FORECASTING

Company Forecasting

Company Forecasting

The attached data shows monthly warehouse store sales (in million dollars) in the USA from Jan. 1992 to Jun. 2004, inclusive. Obtain a time plot of the data.

The Chart that depicts the time plot data of the Warehouse's sales has been done in the attached excel sheet.

Choose, giving your reasons, which exponential smoothing method is likely to be the most suitable for producing forecasts with this data set. Using the mean absolute percentage error (MAPE) as the measure of accuracy, obtain, by trial and error, optimum smoothing constants (up to two decimal places) for this set of data.

For this Data set, triple exponential smoothing would be required (Winter. H, 1960). After observing the time series plot chart, we can conclude that there is a sudden rise in sales revenue during a certain period in every year. Hence these sudden rises in sales depict a season trend for the warehouse's services (Yeung., 2009, pp.n.d). Since the sales revenue of this ware house is seasonal, then triple exponential smoothing will have to be used in order forecast future sales (Winters, 1960). Triple Exponential smoothing considers the weight of seasonal impacts into the forecasting predictions. Using the mean absolute percentage, the best constant was 0.15 with a mean absolute percentage error of only 1.8%

4) Suppose you are in the management services branch of a USA warehouse superstore and that you have just developed the method in part (b), and produced the forecasts in part (c).The actual warehouse store sales (in million dollars) for the periods forecasted in part(c) was as follows;

Jul-04

Aug-04

Sep-04

Oct-04

Nov-04

Dec-04

Jan-05

Feb-05

20549

20477

19108

21013

22486

28274

19813

19691

Write a short report to your head of branch, commenting on the exercise. You should include reference to the accuracy of the forecasts, and the likely accuracy of the method for producing forecasts up to 8 months ahead on a regular basis; potential problems/advantages of the method for producing such forecasts, and how the forecasts might be integrated into the planning operations of your firm.

Note to Branch Manager

Data of past sales figures have been used to calculate the forecasts of the company. By observing the data we have concluded that since the company is experiencing seasonal demand for its product, we will be using a triple exponential smoothing curve. A smoothing constant was calculated by trial and error. Based on the mean absolute percentage error the best smoothing constant that showed the least amount of error was a smoothing constant of 0.15. This smoothing constant showed an error percentage of only 1.8%. By using this smoothing constant we were able to understand the trend through which our sales were rising (Eumetcal, 2012). With the help of the triple exponential smoothing curve, we were able to develop seasonal indices and thus smooth out the variations provided by the seasonal demand. Using the data, we have found out that the sales of the next few months will be as follows.

Jul-04

Aug-04

Sep-04

Oct-04

Nov-04

Dec-04

Jan-05

Feb-05

20549

20477

19108

21013

22486

28274

19813

19691

By using these numbers it mind, we need to allocate all the necessary resources in our entire supply chain to ensure that the expected ...