Forecast Simulation
Forecast simulation is a forecasting technique where several forecast methods are compared to each other. The method which most closely reflects the actual outcome is then selected to use when calculating the base forecast for the next period. This selection is made each time forecasting is done. The information from the forecast method selected in this way is then saved.
Description
This is used as an alternative to the four forecast formulas available for calculating base forecasts.
The following five forecast formulas are used for standard forecast simulation.
| 10 | = | The forecast for the next period is calculated as the mean actual demand from the three previous periods. |
| 11 | = | The forecast for the next period is calculated as the mean actual demand from the three previous periods, but increased or decreased by the expected trend for the item group. |
| 12 | = | The forecast for the next period is calculated as the mean actual demand from the three previous periods multiplied by the ratio of actual demand for these periods and actual demand for the same three periods the previous year. |
| 13 | = | The forecast for the next period is calculated as the mean actual demand from the previous year, for the period immediately before, the same period and the period immediately after. |
| 14 | = | The forecast for the next period is calculated as the mean actual demand from the previous year, for the period immediately before, the same period and the period immediately after, increased or decreased by the expected trend for the item group. |
These formulas are automatically assigned to forecast methods F1, F2, F3, F4 and F5. Other necessary forecast parameters can be defined in each of these methods. Forecast methods F2, F3, F4 and F5 are automatically created for method F1. In this way, when forecast method F1 is selected, the system automatically creates forecast simulation. In addition, different forecast formulas based on the four formulas in the system can be assigned to F1. Additional forecast simulation methods can also be created manually.
Assume the following demand history for an item:
| Oct. 94 | Nov. 94 | Dec. 94 | Jan. 95 | Feb. 95 |
| 90 | 97 | 95 | 98 | 98 |
| Oct. 95 | Nov. 95 | Dec. 95 | Jan. 96 |
| 101 | 104 | 98 | 101 |
The trend for the item group for January 96: 2
The forecast for the item is simulated using the five standard simulation formulas included in the forecast method. The base forecast for January 96 is as follows for each forecast formula.
| 10: | (101 + 104 + 98) / 3 = 101 |
| 11: | 101 + 2 = 103 |
| 12: | 101 * (101 + 104 + 98) / (95 + 97 + 90) = 108 |
| 13: | (95 + 98 + 98) / 3 = 97 |
| 14: | 98 + 2 = 100 |
Since the base forecast using formula 10 is closest to the actual demand for January, this is chosen to forecast demand for February.
The base forecast for February is: (104 + 98 + 101) / 3 = 101.