Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
A forecast is the prediction of a future situation. Aim of demand forecasting is to reduce risk and in planning for firms long term growth. It start with macro economic forecast. Demand and sales of most goods and services are strongly affected by business conditions e.g. sales of automobiles, new houses etc.
DEMAND FORECASTING
A forecast is the prediction of a future situation. Aim of demand forecasting is to reduce risk and in planning for firms long term growth. It start with macro economic forecast. Demand and sales of most goods and services are strongly affected by business conditions e.g. sales of automobiles, new houses etc.
FORECASTING STEPS
A Identification of objective B determining the nature of goods C selecting a proper method of forecasting D Interpretation of results
METHODS OF FORECASTING
Fundamentally two approaches A SHORTRUN- to obtain information about the intentions of consumers by means of market research, survay, economic intelligence etc. B LONGRUN- to use past exprience as a guide and by exploiting past trends, to estimate the level of future demand.
METHODS OF FORECASTING
NEW PRODUCT--- SURVEY METHOD used because no historical data is available. EXPERT OPINION SURVEY METHOD Obtaing information from group of experts regarding future technological stares. Speedy and less costly. DELPHI TECHNIQUE Panel members are asked by letters to give their predictions. They got information throgh post & sent outcome. Those who dissent are invited to give reasons or else modify their forecasts. Process repeated and final range of outcome is regarded as probabilistic forecast.
B SAMPLE SURVEY
Probable demand expressed by each selected unit is summed up. Then total sample demand x by the ratio of number of consuming units in the population.Give good results for new product. ADVANTAGE less tedious less costly less data errors
C END USE METHOD Demand surveys of industries using this product as an intermediate product. Demand for the final product is the end use demand. Intermediate product may have many end use. Domestic and international market demand. The demands for final consumption and exports net of imports are estimated through some other forecasting methods and its demanded for intermediate use estimated through a survey of its users industries. Such a method is feasible for national planning organisations and not for industry. ADVANTAGES 1. Provides sectorwise demand forecasting 2. Does not require historical data 3. If numbers of end users is limited it will be convenient to use this method. DISADVANTAGES It required every industry to furnish its future plan,individual industry will have to rely on some other method to estimate future demand, only intermediate demand can be predicted.
STATISTICAL METHOD
A TREND METHOD Based on analysis of past sales pattern. These methods dispense with the need for costly market research because necessary information is often already available. Time series data- main 5 techniques as : 1 FITTING A TREND LINE BY OBSERVATION
pattren trend
TREND THROUGH LEAST SQUARE METHOD Mathematical Sales = a + b (year ) or S = a + bt Where a & b have been calculated from past data & t is the year number for which forecast is to be made. e.g. The sale record of company x reveal the following Year 1970 1972 1974 1976 1978 1980 Sales in crores 30 40 45 50 48 57 Estimate sales for the year 1982 &1983. SOLUTION To find values of a & b we will have ti solve the normal equation S = Na + b T ST = a T + b T 2 IN THE TABLE BELOW WE FIND S, ST , T, T2
YEAR NUMBER 1 3 5 7 9 11 36
SXT 1 9 25 49 81 121
T2
270 = 6a + 36 b 1784 =36a +266b solving these equations for a & b we get a = 30.76 b = 2.34 Thus the trend equation becomes S = 30.76 + 2.34 t years 1982 &1983 take on the years number 13 &14. By substituting these values for t we get Rs. 61.18 & 63.52 crores. Trend method is popular because it is simple and gives good forecasting more over it does not require knowledge of economic theory and market.
MAIN DRAWBACK Past rate of change in variables continue in future.not appropriate for short term, can not explain turning point of business cycle.
BAROMATRIC TECHNIQUES
It is based on that the future can be predicted for certain events occuring in the present. It involve statistical indicators. Time series provide indications of change in economy or specific industry. These are termed as the barometer for market change. ( 1 ) Leading indicators consists of indicators which move up or down ahead of some series e.g. ( I) index of net business ( capital ) formation ( ii) new orders of durable goods (iii) new building permits ( reflect future market change ). ( 2 ) Coincidental indicators which coincide with or fall behind general economic activity or market trends. e.g. number of employees in non agriculture sector. (3) Lagging indicators-Those indicators which follow a change after sometime e.g. are manufacturers stock level and cosumer credit outstanding. The problem of choice may arise because some of the indicators appear in more than one class.so it is advisable to rely on just one indicator.This lead to uses the diffusion index.
DIFFUSION INDEX
It cops with problem of differing signals given by the indicators. A diffusion index gives the % of rising indicators. In calculating a diffusion index for a group of indicators scores alloted are1 to rising series, to constant series and zero to falling series. If 3 out of 6 indicators are moving up in lagging series the index shall be 50%. It is for short term forecast. QS What is delphi method? What is the use of this method? How baromatric leadership is achieve discuss in light of perfect competition? What are the indicators available in our economy? Present in the class.
REGRESSION METHOD
Regression anlysis denotes method by which the relationship between quantity demanded and one or more independent variables (like income , price ,advertisement ) is estimated. Simple regression analysis is used when the quantity demanded is estimated a function of a single independent variable, such as price. Multiple regression analysis is used to estimate demanded as a function of two or more variables simultaneously.
Like trend fitting method above equation can be estimated by using least square method. The parameter a and b can be estimated by solving the following two linier equations: Y1 = na + bX1 --------------------- 1 XiYi = Xia + bXi ------------------------ 2 calculation of terms in linier equations YEAR POPULATION X SUGER CONSUMED Y X2 XY 1985-86 10 40 100 400 1986-87 12 15 144 600 1987-88 15 60 225 900 1988-89 20 70 400 1400 1989-90 25 80 625 2000 1990-91 30 90 900 2700 1991-92 40 100 1600 4000 n =7 x1 = 152 Y1 = 490 x2 =3994 xy = 12,000
by substituting value in equation 1 and 2 we get 490 = 7a + 152 b -------------------- 3 12,000 = 152a + 3994b ----------------- - 4 by solving equation 3 and 4 we get a = 27.42 b = 1.96 by substituting values for a and b in equation Y = a + bx Y = 27.44 = 1.96 X Given the regression equation, the demand for suger for 1994-95 can be easily projected if population for 1994-95 is known. Suppose population is projected 70 million, the demand for suger in 1994-95 may be estimated as Y = 27.44 + 1.96 (70) =164,640tonnes.
MULTIVARIATE REGRESSION Function of many variables 1 to specify the variables that are supposed to explain the variatios in the demand. The explainatory variables are chosen from the determinants of demand. 2 to collect time series data on the independent variables. 3 to estimate the parametersin the chosen equatios with the help of statistical techniques. MULTIPLICATIVE DEMAND FUNCTION Qx = a Pbx Yc Pd Qx = Quantity demanded for x, Px = price of commodity, y = cosumer income, A = advertisement expences, a is constant, and bcd are parameters expressing the relationship between demand and Px, y, Py, and A.
SIMULTANEOUSLY EQUATION METHOD also known as complete systmatic approach. It involves simultaneously cosiderations of all variables, as it is believed that every variable influences the other variable influences the other variable in an economic decision environment, so here the set of equations equals the number of variables. Mathematical & statistical 1 First step-Develop a complete model and specify the behavioural assumptions regarding the variable included in this model. The variable included in models are called a endogeneous variables b exogeneous variables Endogeneous variables that are determined with in the model as dependent variable Exogeneous variable determined outside model e.g. Govt. tax rate 2 Data collection on both above variable which is hardly available
3 the model is solved for each endogeneous vriable in term of exogeneous variable into the equations, the objective value is calculated and prediction made. e.g. consider a simple macro economic model Yt = Ct + It + Gt + Xt ----------------- 1 Yt Gross National Product, Ct Total cosumption expenditure, It = Gross Private Investment, Gt = Govt. expenditure, Xt = Net Exports, (X M ) where M = Imorts and subscript t represents a given time unit eq.1 is an identity , which can be explained with a system of simultaneously equations. Suppose in eq. 1 Ct = a + b Yt -------------------- 2 It = 20 --------------------------- 3 Gt = 10 --------------------------- 4 Xt = 5-----------------------------5 In the above system of equations, Yt & Ct are endogeneous variables, It, Gt, Xt are exogeneous variable. Eq2 is a regression equation that has to be eliminated. EQ 3 4 5 shows the value of exogeneous variable determined outside model. Suppose we want to predict value of Yt &Ct simultaneously. Suppose also we estimate the eq 2, we get Ct = 100 + 0.75 Yt Now using the equation system, we may determine the value of Yt as Yt = Ct + 20 +10 +5 = Ct +35 since Ct = 100 + 0.75 Yt by substitution we get Yt 0.75 Yt = 100 +35 0.25 Yt = 135 Yt = 135/ 0.25 = 540 now we can calculate Ct easily.
2 MOVING AVERAGE MODEL Model estimates Yt in relation to residuals e1 of the previous years. The generl form of model is: Yt = m +b1 et-1 + b2 et-2---------bp et-p + et where M is mean of the stationary time series. et-1, et-2, et-p residuals, the random components of Y in t-1, t-2, --------t-p periods. 3 AUTO REGRESSION ( MOVING AVERAGE MODEL ) After moving Average model is estimated, it is combined with auto regression model to form the final form of the BOX JENKINS MODEL as below: Y1 = a1 Yt-1 + a2 Y t-2 +------------ +an Yt-n + b1 et-1 +b2 et-2 +-----+ bp et-p + e1 Sophisticated and complicated method require computer.
MOVING AVERAGE METHOD This simple method assumed that demand in a future year equals the average of demand in the past years as below: Dt = 1/N (Xt-1 + Xt-2----------Xt-n) Sales in previous years N = number of preceding year.
Questions
1)You are given the following data:
X 3 14 Y 14 6
8 20
8
6
10
10
13
12
13
12
13
14
Estimate the regression equation Y = a + b X 2) Why is demand forecasting essential? Is demand forecasting equally important for small and big, and old and new business ventures? ( DISCUSS IN THE CLASS ) 3) Demand forecasting is must for the business?