DEMAND FORECASTING

Demand forecasting means estimation of the demand for the goods in forecast period. Demand forecasts may be attempted not only for a total market but also for market segments like domestic demand and foreign demand.

Steps involved in forecasting:

1. Identify and clearly state the objective forecasting.

2. Select appropriate method for forecasting depending upon the objective different tools will be used.

3. Identify the variables affecting the demand for the product.

4. Gather the relevant data to represent the variables.

5. Determine the most probable relationship between the depended and independent variables.

6. Prepare the forecast and interpret the results.

Method of demand forecasting:

There are several methods of demand forecasting basically for 3 reasons

a) No method is perfect and no method is useless.

b) No method is best under all circumstances.

c) The best method may not be available in a particular situation due to constraints from data and resources.

Survey method and statistical method:

Surveys are conducted about the intentions of consumer's, opinion of expert forecast on demand, opinion about market through analysis. Surveys are of two kinds census & sample. Survey methods are suitable for short term forecasts and new product.

Statistical method are common long term forecasts and for forecasting demand for well established Products. Under statistical method historical data extrapolated or analysed through economy models and through them forecast are squeezed out.

Consumer's intention survey method:

Under this method demand forecasts are attempted through a survey of consumer's intention about what they are planning to buy in the Forthcoming time period which is usually within a year. The consumers survey may be done with the help of either a complete survey of all the consumers in the market or by selecting a few out of the relevant population.

Experts opinion survey method:

In this method the experts in the field are asked to provide their own estimates of likely sales. Experts may include managers and executives directly involved in the market, dealers, distributors, suppliers etc.

Delphi Technique:

Delphi Technique may be a formalized forecasting tool of opinion pool or expert opinion. This method consists of an attempt to arrive at a concensus in an uncertain area by questioning a group of experts. Once the members give their opinion the leader of the exercise provide each expert with the responses of the others including their reasons. Moreover this technique comes to a conclusion about the future forecast in a very short time.

Market experiments survey method:

Market experiment method has two versions actual and simulated. Under the actual experiment shops are opened in different localities and then consumers reaction are observed and recorded.

Market simulation method is also called as consumer clinic or laboratory experiment technique. This method involves providing token money to a set of consumers and asking them to shop around in a simulated market. The price of various goods their quality packing etc. Vary during the experiments to observe consumer's reaction to such changes.

Trend projection or Mechanical extrapolation techniques:

In trend projection historical data can be used to predict the future. This method is based on the past sales pattern. Long term trends generally exist because some of the important underlying factors like the population , competition, income etc move steadily and therefore produce only a gradual change over time.

Smoothing method:

If the variables under forecast doesn't follow any specific trend the trend method is inappropriate. There are two versions of the smoothing method

1) Simple smoothing

2) Weighted smoothing

In simple smoothing simple average of the specific number of observation is taken and in weighted smoothing a weighted average is taken out. The weighted smoothing is considered as a preferred smoothing and the weights are assigned in the descending order as one goes from the current observation to the past ones. When the weights are assigned this way the weighted smoothing method becomes the geometrical or exponential smoothing method. There are single simple, double simple, single exponential, double exponential and triple exponential smoothing methods are used. Under all these methods forecasts are obtained through two steps

a) Obtain the specific smoothen series from the observed time series and

b) Obtain the desired forecast from the specific smoothen series.

ARIMA Method:

The Auto Regressive Integrated Moving Average (ARIMA) method has been given by box and Jenkins. So it is also called Box-Jenkin method. The method combines the moving average and auto regressive techniques and thus is considered as the most sophisticated statistical method of forecasting. This method is primarily for short term forecasting.

Leading Indicator method:

This method involves three steps

a) Identification of the leading Indicator for the variable under forecasting.

b) Estimation of the relationship between the variable under forecasting and leading Indicator.

The leading Indicator method of forecasting is also called the Barometric method.

Regression method:

This method employs both the principles of economic theory and appropriate statistical methods of estimation in forecasting demand or any other variable. It requires historical data on the variable under forecasting and it's determinants. Under this method forecasts are obtained through the following steps

* Identification of casual variables.

* Collection of historical data on variable under forecasting.

* Selection of appropriate functional form for the demand function and estimation of the function.

* The derivation of forecasts for the independent variable in the function.

* Derivation of forecasts for the variable under forecasting.

Simultaneous equation method:

Simultaneous equation method is the most sophisticated econometric method of forecasting. Simultaneous equation method overcome the major problem of the regression method ie. Forecast for the independent variables.

Factors affecting Demand forecasting:

a) Business cycle

b) Customer's plan

c) Products life cycle

d) Competitors efforts and price

e) Credit policy

f) Sales effort

g) Advertising.