Demand forecasting

Demand Forecasting

Demand forecasting• A forecast is a prediction or estimation of future situation under given conditions

•Demand forecasting is different from demand estimation in the sense that forecasting predicts about future trends of sales while estimation tries to find out expected present sales level.

Demand forecasting continue…..• Passive forecast: where prediction about future is based on assumption that firm does not change the course of its action

• Active forecast: where forecasting is done under the condition of likely future changes in the actions by a firm

Purpose of forecasting• Short run forecast: seasonal patterns are more important. It helps in preparing sales policy, price policy, production planning to avoid under and over stock conditions.

• Long run forecast: it is helpful for capital planning.

Methods of forecastingMethods of forecasting

Opinion Polling method Statistical methodconsumer’s survey fitting trend line methodsales force opinion least square methodexpert’s opinion moving average

exponential smoothing

Opinion survey/consumer’s survey• Relatively simple and practicable method for forecasting demand of new products•Opinions are collected from prospective buyers regarding their future consumption• Sampling technique is used to survey customers• From sample it is possible to forecast demand of targeted population

Expert’s opinion• In this method expert’s opinion is sought on the future demand for product• It is biased and subjective• The accuracy of predicted demand depends on skill, expertise and experience of person making forecast•Method is useful for forecasting demand of established product

Sales force opinion • Expected sales is estimated by distributors survey through questionnaire or can be requested from retail outlet• Company’s sales force can also give estimation of future demand•Many company heavily rely on judgment made by their sales personnel• But this judgment may suffer from over optimism or over pessimism

Delphi technique

• It can make more realistic forecast• A panel of experts are asked sequential question and from responses new questionnaire is produced.•Opinions are collected from experts to arrive at reliable results• Each questionnaire demands a detailed opinion from each expert and then these opinions are summarized to get result

Time series methods

• Time series refers to past data arranged in chronological order as a dependent variable and time as independent variable for ex.

• This is called time series. This method does not study factors affecting demand. In this method all factors that affect demand are grouped into one factor ‘Time’ and demand is expressed as a series of data with respect to time

Year 1994 1995 1996 1997

Demand 20 30 40 58

Fitting Trend Line by observation method• The given time series data are plotted on a graph paper by taking time on X-axis and the other variable on Y-axis.• A smooth line or curve, drawn through the plotted points would represent the trend of the given data.

Least Square method (Regression Analysis)• In regression analysis relation between dependent variable (y) and independent variable (x) can be expressed by equation:

Y=a+bX

Moving Average• Past data can have fluctuations because of seasonal variation and random variation

• Averaging the demand for previous period is going to hide the trend

•MA consists series of arithmetic means calculated from overlapping groups of successive elements of time series.

• The period of moving average should be carefully selected

•Wrongly selected period will distort data.

• Longer the period of M.A. greater is the smoothing effect

Moving Average continue…• Each moving average is based on values covering a fixed time interval, called ‘period of moving average’ and is shown against centre of the period.• For the time series values Y1, Y2, Y3,… the moving average for period n is given as follows:

Moving average continue….• 1st value of M.A.= 1/n (Y1+Y2+Y3+…+Yn)• 2nd value of M.A.= 1/n(Y2+Y3+Y4+…+Yn+1)• 3rd value of M.A. = 1/n (Y3+Y4+Y5+…+Yn+2)And so on…..

Continue….•When period of M.A. is odd the successive values of the moving average are placed against the middle period. • For ex. If n=7 then first moving average is placed against 4th value, the second moving average is placed against 5th value and so on.• If the period of M.A. is even then centering method is used

Exponential Smoothing method• Very popular approach for short term forecasting•Method determines value by computing exponential weighted system• The weights are so assigned that w lies between 0 and 1• The rate of smoothness depends on value of w.

• The smoothing scheme begins by setting smoothened value equal to observed value for 1st period that means

S1=Y1

• And for succeeding time period t, smoothened value St is found by equation

St= w.Yt + (1-w).St-1

Where St = current smoothened valueYt = current observed valueSt-1 = previous smoothened value