, 0017654096289



2. Chapter:

Demand forecasting:

Meaning, objectives and determinants of demand forecasting: the concept of certainty, risk and uncertainty- factors affecting demand-methods-survey method and statistical methods-semi average, moving average and least square methods (with emphasis on problems)

Demand Forecasting:

Meaning:

A forecast is a prediction about the future event which is likely to happen under given conditions. Good production and sales planning require forecast of the business conditions and of their relationship to demand. Forecast is essential to minimize the uncertainties of the unknown future. The more realistic the forecast, more effective decisions can be taken for the future. Usually as a first step, the firm needs to make a sales forecast and then later on move to other areas of forecasting and planning.

Scope of demand forecasting:

Before discussing the methods of forecasting, it is necessary to find out the specific purpose of the forecast. A number of conditions can be discussed. They are:

1. Period of forecast:

Firms have to forecast for both short and long periods. The first step therefore, is to decide about the length or period for which the forecast is done. Short period may mean a maximum of one year. While long period could be anything between five years to twenty years. Some economists also make a third classification of medium term, which is in-between short and long period is simple. Because, it is only the seasonal factors which influence short period forecasting. Long period forecast is more complicated because it depends on the nature of the commodity. Whether it is a consumer good, durable good, intermediate good etc. Long period forecast becomes necessary in the case of goods like iron and steel, petroleum, pharmaceuticals which require heavy investment and therefore production has to be planned carefully.

2. Levels of forecasting:

Forecasting may be undertaken at different levels like:

a. Macro economic forecasting:

This is concerned with business conditions, covering the entire economy. Some indicators like national income, total production, existing prices, are taken into consideration, while measuring the business conditions. In other words, these are the assumptions on which demand forecast is based.

b. Industry or market demand forecasting:

These forecasts indicate in which direction the entire industry will be moving. For example: manufacture of one brand of T.V. say, BPL would want to known in what way the T.V. industry is likely to behave so as to decide about the way his firm should plan for the future. These forecasts based on consumer’s intensions survey are generally supplied by the trade association. Such forecasts are used by the firms to compare industry sales with their own and thus determine their future market share.

c. Firm or company demand forecasting:

In the case of big firm for example: Godrej, would like to find out and forecast the demand for its own products independent of the rest of the firms in the industry. It suggests whether the firm is capable of maintaining or improving its share in the industry.

d. Product-line forecasting:

If a firm producing different types of goods, then, productline forecasting will help a firm to decide on its priority allocation. This forecast is essential because every firm faces the problem of limited resources. For Example: A firm like Hindustan Lever will have to decide whether to produce more of soaps or detergents (washing powder).

3. Nature of production:

Different methods of forecasting have to be adopted for different types of products. An old product enjoys better advantages than new products, which are introduced for the first time. Therefore, it is comparatively easy to forecast the demand for old products, while the new one faces uncertainity as far as its demand is concerned. Similarly, forecasting will be different for different types of goods like capital goods, consumer durable goods and non-durable goods.

4. Market situation:

The changing conditions in the market also, play an important role in influencing the demand for a product. If for example: a product has number of substitutes, then, the producer is always facing a risk and may find it difficult to forecast the demand. On the other hand, if one has monopoly, the degree of risk is less and he can easily predict the future for his commodity.

Objectives of demand forecasting:

The objectives of demand forecasting can be divided into short-run objectives and long-run objectives.

I. Short-run objectives:

a. To fix an appropriate level of production:

The firm has to produce according to a plan. If greater demand is expected at a future date, the production level has to be geared up to that level. This can be done only through demand estimation.

b. To reduce costs:

If estimation is accurate a number of costs, can be reduced. For example: costs involved in the purchase of raw-materials, wages to casual labour can be cut down through proper demand forecasting.

c. To fix prices:

Forecasting helps to determine prices according to the existing market conditions, what should be the price when demand is high or by how much price should be lower when demand is low, can be easily understood through accurate forecasting of demand.

d. To fix sales target:

Correct sales targets can be fixed when demand forecasting is definite, higher sales targets can be fixed if estimated demand is high. Incentives can be worked out in advance to boost up sales if necessary.

e. Sales promotion:

Advertising of the product through different media can be planned well in advance if demand estimation is available.

f. Financial planning:

Arrangements for additional finances can be made in case demand is likely to go up in the short period. In the absence of estimation it becomes extremely difficult for firms to arrange for funds at short notice.

II. Long-term objectives:

a. Production planning:

If demand is likely to increase in the years to come, proper expansion of the unit will have to be undertaken. In many cases new units may have to be put-up. On the other hand, if the future demand is likely to fall, gradual reduction in the strength of the workers, amount of investment, etc. will have to be under taken. In many cases new units may have to be put up. On the other hand, if the future demand is likely to fall, gradual reduction in the strength of the workers, amount of investment etc, will have to be under-taken.

b. Planning for personnel:

Man-power planning is very essential for the success of a firm. Long term projection of demand will help the business unit to plan properly the man-power required for the unit. In many large units, the unit itself may train up the necessary personnel required.

c. Financial planning:

Large firms require considerable amount of money for investment purposes. Such firms will have to think of devices like shares and debentures to meet their financial requirements. Therefore, large investments need systematic planning and long-term projection which will give an idea of the financial requirements of the business unit.

Steps involved in demand forecasting:

1. Identification of objectives:

It is essential to be clear about what one wants to get from the forecast. For example: the purpose of the demand forecast could be the estimate of either the quantity of demand the price to be fixed, sales planning etc.

2. Determing the nature of the commodity under consideration:

The next step is to find out the category to which the commodity belongs. The classes being capital goods, durable goods etc. each of them has their own different demand pattern.

3. Selection of the method of forecasting:

Choosing a proper method for forecasting the demand will depend upon a number of factors like types of commodities, period of forecast, type of data available and the objective of the forecast.

4. Interpretation of results:

The most important step is the analyzing of the result obtained after the forecast is made. Most of the time, the forecast has to be supported by factors like government policy, general business conditions, international economic and political situation over which the firm does not exercise any control, further, it is necessary to revise forecast when circumstances change, because forecast are generally made on the assumption of continuation of past events.

Criteria for the choice of a good forecasting method:

There are six ways of choosing the best forecasting method. They are:

1. Accuracy:

This is the most important test that a demand forecasting method has to face. Precise estimation may not be possible due to changes in purchasing power. Taste and preferences etc. but, estimation must be accurate atleast to satisfactory level. Marginal variations may be accepted.

2. Plausibility:

The method used should be understood easily by the business firm or a manager. In recent years some of the mathematical techniques which have been developed are so complicated that they can be understood by very few people. Very often, the predictions made by firms on the basis of their experience are more accurate than any of the mathematical methods. In this case, the former method is more plausible than the latter.

3. Durability:

The power of the estimation should be strong, with the present demand conditions it should be possible to predict for the future. Similarly, it should be possible to use the same function to move back and state the previous demand. The prediction made should hold well for sometime.

4. Flexibility:

With change in time, the method used should also allow for necessary changes in the determination of variables. For example: if we say that demand is an exclusive function of price, then the future demand forecast must allow for a possible change in price. The forecast therefore, must be flexible one to accommodate any likely changes in variables affecting the demand.

5. Availability:

The forecast should be such that it makes full use of the available data or depending on the data available, a suitable method could be chosen. Data can be collected either through surveys or other secondary sources and must make use of all the information available.

6. Economy:

The method selected should not be an expensive one. In the case of the smaller firms specially, it is not possible for such firms to spend large amounts on forecasting because their budget itself would be limited. However, the method chosen should not force the business firm to compromise on the accuracy of the results obtained.

Methods of demand forecasting:

At the outset it should be made clear that there is no easy or simple method which will help the firm to predict the uncertainties of the future. There are various forecasting techniques, differing in their accuracy; the problem is, to choose the most efficient method given the objective of the forecast and the nature of information.

Mathematical techniques are essential in classifying relationships, but, they cannot replace sound judgment. In the same way through judgment is one of the requirements of a good forecast it is incomplete without analysis. Therefore, forecasting must depend both on judgment and on analysis. The methods used for demand forecasting are:

I. Opinion polling method

II. Mechanical Extra polling Method

III. Barometric method

IV. Statistical methods

I. Opinion polling method:

These methods are generally used for short period forecasting. The different types of opinion polling methods are:

Complete enumeration method:

Under this method all the consumers of the product are interviewed based on which forecast is made.  As first hand information is collected, this method is free from bias.  However, this method is impractical as the consumers are numerous and scattered.

a. Opinion surveys :( surveys of buyers’ intension):

Under this method consumers are contacted personally to find out their future buying plans. This method has the advantage of first hand information, there are certain disadvantages also, contacting a large number of consumers scattered in different places may be difficult. Secondly, it is costly and time consuming method. Thirdly, consumers may be unwilling to discuss future buying habits because of personal privacy or commercial secrecy. Fourthly, the consumer’s may miss judge their future purchases or change their plans due to unexpected changes in future conditions.

Cost of such a survey can be reduced if many firms take-up joint survey programmes. However, this method is not a reliable one because accuracy is less and no firm can base its forecast fully on such a survey.

b. Sales force opinion method (collective method):

Instead of conducting a survey of consumer’s opinion, a survey is carried out on the opinion of the sales persons because they are closet to the market. This is a cheap and easy method, because, no complicated statistical measurement is involved. This method is generally used to forecast demand for new products. It also has the advantage of first hand information obtained from sales people.

But there is a major drawback in this method. Most sales people would either over estimate or underestimate the changes in future market conditions. This results in either exaggeration or deflation of future estimates, to overcome, this problem to some extent, a record of each sales persons achievements and success in forecasting would be maintained and adjustments made accordingly.

c. Panel of experts method:

This method involves collecting the opinion of a group of experts or specialists outside the firm. It has the advantages of speed as well as being inexpensive. This method is generally used under those conditions where data may not be available in plenty.

One example of this method is the ‘Delphi Technique’ under this techniques, some experts are asked by letters to give their predictions of the occurrence of specified events. Each person is ignorant about the responses of the others because of postal anonymity is maintained. The members therefore can express their opinions without any personal inhibitions. The responses of all the members are then discussed by the firm and depending on the result forecasts are made, members are also informed of the outcome of such results and any member who disagrees will be asked to specify his reasons for dissent. Otherwise they are asked to modify their forecast.

d. Test marketing method:

This method is useful for predicting the sales of new products or to find out the potential of old products in new areas. Under this method, a test area is selected which can be regarded as a representative portion of the entire market. Then the product is launched in the national market, package designs, sales, TV, and press support, price etc must be selected which the national market in view. If the product is successful in the test area, then forecast can be made, that the product will be successful regionally or nationally.

But this method is time consuming as well as expensive. The product must be produced in a significant amount but only a portion of it is sold in the test area. A full scale marketing effort is involved which means considerable amount of money is spent on sales promotion. The test must be continued long enough to permit consumers’ repurchase cycle to operate. It is difficult to select a test area, which is typical of the entire market. The test area must be similar to the national market in terms of occupational groupings, age-sex compositions, income-levels, advertising media availability etc. It is difficult to find such a test area which represents the entire market. Very often, when one firm is test marketing its product, its rival may introduce a cheap imitation of the same product and sell it nationally without incurring the cost of test-marketing.

e. Sample survey method:

When a commodity enjoys huge number of consumers, complete surveys cannot be undertaken. In such an event, the next best survey method can be used that is the sample survey method. Under this method the forecaster selects a few samples from among the population and interviews them, he may make use of either random sampling technique or stratified sampling technique.

1. Random sampling technique:

Under this method, consumers are selected at random. Normally, three methods are used to select the samples, they are:

i. Lottery method or ticket sampling method:

In this method relevant feature of each consumer is written down on similar cards. Then the card is mixed up and the required number of cards is picked as samples.

ii. Ordinal sampling method:

Under this method, every Kth consumer is chosen. The consumer names are written in alphabetical order and every Kth name is selected.

iii. Random sampling numbers:

Tables of random numbers are used in this method. There are many tables like:

*LHC tippet:

Tracks of computers which give 10,400 four figure numbers composed of 41600 digits.

* M.G.Kendall and Badington Smith:

This gives hundreds, thousands of digits grouped in two and fours.

*R.A.Fisher and Yates:

Which is a table generally used for agricultural and medical research.

* A million random digits published by the rand co-operation, which is usually used in consumption demand forecasting.

2. Stratified sampling methods:

In this method, the consumers are divided into various categories and then the sample is chosen. For example: consumers may be classified into urban and rural or high income or low income and then the sample is chosen.

For any sampling method, the sample should be chosen carefully and the questioner should be of the right type, stratified sampling is widely used in forecasting the demand for new products.

f. End-use method:

The sale of a commodity is projected through a survey of its end use. A commodity may be used by several industries as an input or by consumers when it becomes a final product. It may also be either imported or exported. Therefore, a survey of the industries using that commodity, a survey of the consumers using the same commodity imported or exported is carried out. Depending on all these surveys future demand is forecasted.

Advantages of end-use method:

The advantages of end use method are:

1. It is possible to point out exactly the deviations in demand at any time.

2. Future demand can be manipulated because information is obtained from different sectors individually.

g. Controlled experiment method:

Under this method the firm changes the price of its commodity in different markets and observes the sections of the consumers.

For example:

A firm may reduce the price in one market and find out the change in demand as compared to the original demand or it may choose one geographical area and advertise the product intensively in order to find out how demand is affected.

However, this method suffers from the following drawbacks.

1. What holds good for one market may not always hold good for another.

2. Price is not the only determinant of demand.

Inspite of the drawbacks some business firms makes use of this method to find out the popularity of a product. If the product is popular in one region it can be sold throughout the country through price changes and better sales promotion.

II. Mechanical extra pollution method or trend projection method:

Under this method once the demand function for a commodity is determined at a given point of time it becomes easy to determine and forecast the future demand. These methods are generally based on past sales patterns. It dispenses with the need for costly market research because the required information is already available in the company records itself. The past information is considered useful for predicting future sales. Many algebraic methods are used to forecast the demand like:

a. Parabolic curve method

b. Neutons method

c. Binomial expansion method

Because of the use of mathematics these methods provide more accurate results than the survey methods, provided accurate statistical data is available about sales patterns.

III. Barometric technique or leading indicator method or economic barometers:

Many variables lead other variables in business. For example: if bank rate changes interest rate will also change. Similarly, if loans for house building are easily available, then demand for bricks, cement, wood etc will also rise. The movement of all connected variables can be easily predicted through the measure of the leading variable. The leading indicator is called the barometer.

This is a simple method because once the leading variable is established then it is easy to forecast the demand for other variables. But in some business activities, it is difficult to find out which is the leading variable or the leading indicator may change over a period of time.

IV. Statistical demand analysis:

These methods are generally used to forecast demand for long periods. Some of the methods used are:

1. Index method:

There are different types of indices like purchasing power index, quality index, and sales production index.

Purchasing power index means the capacity of the consumers to buy goods. Purchasing capacity denotes the demand. This index can be calculated if population of the area, share of income of that area and sales percentage of the same area are known.

Example:

Consider the sale of washing machines in Mysore. The following data is give:

1. Percentage of population of Mysore

To all India population { 0.06

2. Percentage of income of Mysoreans

Compared to national income { 0.04

3. Percentage of sales of washing

Machines in Mysore compared to

Total washing machines sold in the

Country during the years { 0.11

___________

0.21

Therefore purchasing power index of Mysore relating to washing machine is:

P.P.I = 0.21/3 = 0.07

2. Quality index (QI):

QI=PPI

_____ x 100

Percentage of population

QI = 0.07

_____ x 100

0.06

QI= 116.6

If the index is greater than 100, it should be inferred that there is market potentiality for washing machines in Mysore.

3. Sales production index (SPI):

SPI = percentage of sales

________________x100

Percentage of population

SPI= 0.11

______x 100

0.06

SPI= 183.33

If the index is greater than 100, it should be concluded that sales potential is favourable.

2. Echelon method forecasting:

Here the forecast are made from macro level. First the demand for the economy as a whole is forecasted, then industry forecasting to firm forecasting to specific geographical region. Data relating to national income, total savings, and consumption are made use of.

3. Holistic method of forecasting:

This method is opposite to the echelon method. Hence forecasting is performed from micro level to macro level. First, the demand of the firm is forecasted, then industry and finally economy level forecasting. The advantage under this method is that, data is more easily available for a firm than for the entire economy.

4. Trend method:

Under this method past data is used to forecast the future demand. Data relating to a commodity over a period of time is called” time-series data”.

Trend method may also be referred to as time series method.

For example: if, the data is observed for over a long period of time, it will be found that there are fluctuations in the demand or sales. The trend method tries to overcome such fluctuations through a method of averages.

Time series data consists of four components namely:

a. Secular trend

b. seasonal variation

c. Cyclical variation

d. Random variation

a. secular trend:

There are long term changes reflecting either a continuous growth or continuous decline over a long period of time.

b. Seasonal variation:

Changes in demand maybe due to changes in seasons. For example: The demand for cotton clothes or woollen clothes, fans or room heaters depends on seasons. Hence the demand for such goods may indicate regular ups and downs over a period of time depending on the season.

c. Cyclical variation:

Due to the movement of the trade cycle, data may also vary. Prosperity, recession, depression and recovery, all record changes in data over a time period. They have their effect on the income of the people thereby affecting the demand.

d. Random variations or erratic variations:

Demand may also be influenced by some random factors. For example: if crops fail due to drought or new substitutes are introduced which will result in demand fluctuations.

There are four methods which are used to estimate the trend. They are:

I. Free- hand method

II. Semi average method

III. Moving average method

IV. Method of least squares

I. Free-hand method:

This is a graphical method where the year is measured along the ‘x’ axis and the demand along the ‘y’ axis. Then the data is plotted on the graph. After that a smooth curve is drawn through the points in order to describe the general long run tendency of the demand, ignoring any short-period fluctuations.

Illustration 1:

The demand data for coffee for six years is given below. Obtain the trend line using the free-hand method.

Year Demand (m.ton)

1984 3

1985 7

1986 2

1987 7

1988 11

1989 12

The demand for coffee is illustrated in the graph. A trend line ’AB’ is obtained on the basis of the data. This line is an approximation of the demand for coffee without taking into account the fluctuations in the demand. If ‘AB’ line is extended further, the demand forecast can be made for future years.

This method is simple and quick method of estimating the demand for a commodity.

II .Semi-average method:

In this method, the given data is divided into 2 equal parts. Then the average of each part is represented as the semi-average. Then the trend line is obtained using the semi-average.

Illustration 1:

Consider the following data which represents the demand for cotton from 1980-85

Year demand

1980 15

1981 20

1982 25

1983 20

1984 30

1985 40

Use the semi-average method to forecast the demand.

Solution:

The given data is divided into 2 equal parts. The total is calculated.

1980 – 1982 [15+20+25]

= 60/3= 20

1983 −1985 [20+30+40]

= 90/3 = 30

III. Moving average:

This is a method by which the demand for a commodity is obtained by taking successive averages of the time series data. Successive moving averages are calculated for the data by omitting every time the first observation and adding the one immediately after the last. This method smoothens out the fluctuations in demand longer the period chosen for the average smoother will the trend. Generally, 3 year, 5 year and 7 year moving averages are computed.

3 year moving average:

Consider the demand for tea compute the three year moving average

Year Demand computation of 3yrM.A. T.V.

1974 40 -----------

1975 30 40+30+20 = 90/3 = 30

1976 20 30+20+40 = 90/3 = 30

1977 40 20+40+60 = 120/3 = 40

1978 60 40+ 60+20 = 120/3 = 40

1979 20 60+20+50 = 130/3 = 43.3

1980 50 -------------

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download