What do you understand by demand forecasting? Explain its method

Ans.. After gathering information about various aspects of the market and demand from primary and secondary sources, an attempt may be made to estimate future demand. Several methods are available for demand forecasting. The important ones are—

(i) Trend projection method

It consists of (i) determining the trend of consumption by analyzing past consumption statistics, and (ii) projecting future consumption by extrapolating the trend. The trend of consumption may be represented by one of the following relationships:

Linear Relationship: Yt = a + bt … (1)

Exponential Relationship: Yt = aebt … (2)

On logarithmic transformation this becomes:

Log Yt = log a + bt

Polynomial Relationship: Yt = a0 + a1t + a2t2 + … + antn … (3)

Cobb Douglas Relationship: Yt = atb … (4)

On logarithmic transformation this becomes:

Log Yt = log a + b log t

In the above equations Yt represents demand for year t, t is the time variable, a, b and aj’s are constants.

Out of the above relationships the most commonly used relationship is-

Yt = a + bt

his relationship may be estimated by using one of the following methods: (i) visual curve fitting method, and (ii) least squares method.

Evaluation— The basic assumption underlying the trend projection method is that the factors which influenced the behaviour of consumption in the past would continue to influence the behaviour of consumption in the future. This hypothesis is sometimes referred to asthe hypothesis of “mutually compensating effects”. Clearly, this is a deterministic hypothesis of questionable validity. Notwithstanding this weakness, the trend projection method is used popularly in practice. Often a starting point in the forecasting exercise, it is likely to be relied upon heavily when no other viable method seems available. The ease with which it can be applied may induce a sense of complacency.

(ii) Consumption level method

Useful for a product which is directly consumed, this method estimates consumption level on the basis of elasticity coefficients, the important ones being the income elasticity of demand and the price elasticity of    demand.

Income elasticity of demand— The income elasticity of demand reflects the responsiveness of demand to variations in income. It is measured as follows:

Q2 – Q1 I1 + I2

E1 = ———— × ———

I2 – I1 Q2 + Q1

15

Where E1 = income elasticity of demand

Q1 = quantity demanded in the base year

Q2 = quantity demanded in the following year

l1 = income level in the base year

l2 = income level in the following year

Example— The following information is available on quantity demanded and income level:

Q1 = 50, Q2 = 55, I1 = 1,000, and I2 = 1,020. The income elasticity of demand is-

55 - 50 1,000 + 1,020

E1 = ——————— × ——————— = 4.81

1,020 – 1,000 55 + 50

The information on income elasticity of demand along with projected income may be used to obtain a demand forecast. To illustrate, suppose the present per capita annual demand for paper is 1 kg and the present per capita annual income is Rs. 1,2000. The income elasticity of demand for paper is 2. The projected per capita annual income three years hence is expected to be 10 per cent higher than what it is now. The projected per capita demand for paper three years hence will be-

Present per 1 + per capital change income elasticity capita income in income level of demand

= (1) (1 + 0.10 x 2) = 1.2 kg.

The aggregate demand projection for paper will simply be-

Projected per capita demand × Projected population

The income elasticity of demand differs from one product to another.Further, for a given product, it tends to vary from one income group to another and from one region to another. Hence, wherever possible,disaggregative analysis should be attempted.

Price elasticity of demand— The price elasticity of demand measures theresponsiveness of demand to variations in price. It is defined as—

Q2 – Q1 P1 + P2

Ep = ———— × ———

P2 – P1 Q2 + Q1

Where, Ep = price elasticity of demand

Q1 = quantity demanded in the base year

Q2 quantity demanded in the following year

P1 = price per unit in the base year

P2 = price per unit in the following year

Example— The following information is available about a certain product:

P1 = Rs. 600, Q1 = 10,000, P2 = Rs. 800, Q2 = 9,000. The price elasticity of demand is:

9000 – 10,000 600 + 800

Ep = ——————— × ——————— = - 0.37

800 - 500 9,000 + 10,000

The price elasticity of demand is a useful tool in demand analysis. The future volume of demand may be estimated on the basis of the price elasticity coefficient and expected price change. The price elasticity

coefficient may also be used to study the impact of variable price that may obtain in future on the economic viability of the project. In using the price elasticity measure, however, the following considerations should be borne in mind: (i) the price elasticity coefficient is applicable to only small

variations. (ii) The price elasticity measure is based on the assumption that the structure and behaviour remain constant.

(iii) End use method

Suitable for estimating the demand for intermediate products, the end use method, also referred to as the consumption coefficient method involves the following steps:

1. Identify the possible uses of the product.

2. Define the consumption coefficient of the product for various uses.

3. Project the output levels for the consuming industries.

4. Derive the demand for the product.

Projected Demand

Consumption

coefficient*

Projected output

in Year X

Projected demand

in Year X

Alpha 2.0 10,000 20,000

Beta 1.2 15,000 18,000

Kappa 0.8 20,000 16,000

Gamma 0.5 30,000 15,000

Total = 69,000 tones

*This is expressed in tones per unit of output of the consuming industry. As is clear from the foregoing discussion, the key inputs required for the application of the end-use method are— (i) projected output levels of consuming industries (units), and (ii) consumption coefficients. It may be difficult to estimate the projected output levels of consuming industries (units). More important, the consumption coefficients may vary from one period to another in the wake of technological changes and improvements in the methods of manufacturing. Hence, the end-use method should be used judiciously.

(iv) Leading Indicator Method

Leading indicators are variables which change ahead of other variables, the lagging variables. Hence, observed changes in leading indicators may be used to predict the changes in lagging variables. For example, the change in the level of urbanization a leading indicator may be used to predict the change in the demand for air conditioners a lagging variable.Two basic steps are involved in using the leading indicator method: (i)First, identify the appropriate leading indicator(s). (ii) Second, establish the relationship between the leading indicator(s) and the variable to be forecast.The principal merit of this method is that it does not require a forecast of an explanatory variable. It, however, is characterized by certain problems.

(i) It may be difficult to find an appropriate leading indicator(s).

(ii) The lead-lag relationship may not remain stable over time. In view of these problems this method has limited use.

(v) Econometric method

An econometric model is a mathematical representation of economic relationship/s derived from economic theory. The primary objective of econometric analysis is to forecast the future behaviour of the economic variables incorporated in the model.

Two types of econometric models are employed: the single equation model and the simultaneous equation model. The single equation model assumes that one variable, the dependent variable (also referred to as the explained variable), is influenced by one or more independent variables (also referred to as the explanatory variables). In other words, one-way causality is postulated. An example of the single equation model is given below:

Dt = a0 + a1Pt + a2Nt

Where, Dt = demand for a certain product in year t

Pt = price for the product in year t

Nt = income in year t

The simultaneous equation model portrays economic relationships in terms of two or more equations. Consider a highly simplified three equation econometric model of Indian economy.

GNPt = Gt + It + Ct … (5)

It = a0 + a1 GNPt … (6)

Ct = b0 + b1 GNPt … (7)

Where GNPt = gross national product for year t

Gt = governmental purchases for year t

It = gross investment for year t

Ct = consumption for year t

In the above model, Eq. (5) is just a definitional equation which says that the gross national product is equal to the sum of government purchases, gross investment and consumption. Eq. (6) postulates that investment is a linear function of gross national product; Eq. (7) posits that consumption is a linear function of gross national product.The construction and use of an econometric model involves four broad steps.

1. Specification— This refers to the expression of an economic relationship in mathematical form. Equation (6), for example, posits that investments is a linear function of gross national product.

2. Estimation— This involves the determination of the parameter values and other statistics by a suitable method. The principal methods of estimation are the least squares method and the maximum likelihood method, the former being the most popular method in practice.

3. Verification— This step is concerned with accepting or rejecting the specification as a reasonable approximation to truth on the basis of the results of estimation and appropriate statistical tests applied to them.

4. Prediction— This involves projection of the value of the explained variable(s).

Evaluation— The econometric method offers certain advantages- (i) The process of econometric analysis sharpens the understanding of complex cause-effect relationships, (ii) the econometric model provides a basis for testing assumptions and for judging how sensitive the results are to changes in assumptions.

The limitations of the econometric method are— (i) it is expensive and data-demanding. (ii) to forecast the behaviour of the dependent variable, one needs the projected values of independent variable (s). The difficulty in obtaining these may be the main limiting factor in employing econometric method for forecasting purposes.

Market penetration for the product— Once a reasonably good handle over the aggregate demand is obtained, the next logical question is: What will be the likely demand for the product of the project under examination? The answer to this question depends on—

1. Aggregate potential supply

2. Nature of competition

3. Consumer preferences

4. Sales promotion efforts

If the aggregate potential domestic supply is likely to be significantly less than the aggregate potential domestic demand, the demand for the product of the project under examination is likely to be very strong, provided liberal imports which may hurt domestic manufacturers are not allowed. The nature of competition and market-sharing arrangement (if any) has a bearing on the demand for the product of the project under examination. Consumer preferences for competing products and the sales promotional efforts of various competitors obviously influence the relative market shares enjoyed by them.