For seasonal products, use the automatic trend calculation to even out the demand forecast according to recent trends. The system calculates trend in a hierarchical order. When you enable automatic trend calculation, the system calculates a trend factor by comparing the recent demand to that of the prior year, and then multiplies the current demand forecast by the factor. Parameters set at the product level override those set at the buy line level. Parameters set at the buy line level override those set at the branch level.
For products with automatic trend calculation enabled, the system does the following:
Calculate the current demand forecast based on the as of date, and note the number of days of sales history it took to meet the minimum days or minimum hits forecast parameters.
Determine the recent demand for the product, starting from the as of date and going back the number of days noted in Step 1.
Determine the product demand for the same period in the prior year, starting from 365 days before the as of date and going back the number of days in Step 1.
Calculate the trend factor by dividing the demand of Step 2 by the demand of Step 3.
If the system-calculated automatic trend percentage is higher or lower than the Max Increase% or Max Decrease%, respectively, then the system defaults to these values.
If the demand for step 3 is 0 (zero), the Auto Trend factor is set to 1 (one) and the Demand Calculation Audit window displays N/A in the Automatic Trend field.
Produce a revised demand forecast by multiplying the current demand determined in Step 1 by the trend factor.
Note: If the product uses non-seasonal forecasting or Auto Trend is set to N, the Demand Calculation Audit window displays N/A in the Automatic Trend field.
For example, the current demand based on the As of Date is 49 days. The recent demand for the product is 192.5 each. The product demand for the same period the previous year is 166.1 each. The system calculates the trend factor as follows: (1 - (166.1 / 192.5)) = 13.71 or 13.71%. For information on auditing demand, see Auditing Demand Calculations.
See Also: