n (iii) Multiply time deviations with the sales of each period individually (XY) and add the product of the column to find (EXY). (iv) To find the trend (Y) this is equal to a +bX. The value of a and b may be determined by either of the following two ways: (a) Direct method. This method is applicable only when X = 0. To make x = 0, it is necessary that the time deviations should be calculated exactly from the mid point of the series. Then, the values of a and b will be calculated as follows: a (average) = ? and b (rate of growth) = X This method is simple and direct. (b) Indirect method. This method is somewhat difficult. This method can be applied in both the cases where Ex has any positive or negative values or EX is not equal to zero. The values of a and b are calculated by solving the following two equations: = n + b = X + b2 By calculating the values of a and b in the above manner, the sales can be forecasted for any future period by applying the formula Y = a +bX. 7. Moving average method: This is another statistical method to calculate the trend through moving averages. It can be calculated as follows: An appropriate period is to be determined for which the moving average is calculated. Whie determining the period for moving averages, the normal cycle time of changes in the values of series should be considered so that short-term fluctuations are eliminated. As far as possible, the period for moving averages should be in odd numbers such as period of 3.5 or 7 years. The period in even numbers will create a problem in centralising the values of averages. The calculated values of moving averages present the basis for determining the expected amount of sale. v1 8. Criteria of a good forecasting method: It cannot be said which method of sales forecasting is the best because everyone has merits and demerits of its own. The suitability of a method depends on various factors such as nature of the product, available time and past records, wealth and energy, degree of accuracy and the forecaster etc. of an enterprise. However, in general, a good forecasting method must possess the following qualifications. 0 Accuracy: Accuracy of the forecasting figures is the life blood of a business because many important plans and programmes, policies andstrategies are prepared and followed on the basis of such estimates. If sales forecasts are wrong, the businessman suffer a big loss. Hence, the method of forecasting to be applied must amount to maximum accuracy. (ii) Simplicity: The method for forecasting should be very simple. If the method is difficult or technical, then there is every possibility of mistake. Some information are collected from outside and that will remain unanswered or inaccurate replies will be received, if the method is difficult. Management must also be able to understand and have confidence in the method. (ii) Economy: The method to be used should be economical taking into account the importance of the accuracy of forecast. Costs must be weighted against the importance of the forcast to the operations of the business. (iv) Availability: The method should be such for which the relevant information may be avaiable immediately with reasonable accuracy. Moreover, the technique must give quick results and useful information to the management. (M) Stability: The data of forecasting should be such wherein the future changes are expected to be minimum and are reliable for future planning for sometime. (vi) Utility: The forecasting technique must be easily understandable and suitable to the management. Illustration 1. From the following time series data of sale project the sales for the next three years. Year Sales 1 000 units 2001 80 2002 90 2003 92 2004 83 2005 94 2006 99 2007 92