The simple moving average method gives equal weight to each of the historic usage figures. The figure for January contributed equally to the forecast for July as the usage figure for June. This does not account for the fact that older figures are less reliable than more recent figures. The weighted average, or exponential smoothing method of forecasting takes into account gradual trends in item usage to provide more accurate forecasting. The method gives greater weight to usage figures recorded in recent months. Table 3 shows an example of a monthly weighted average forecast. Whilst the total values are similar, the weighted forecast smooths out major peaks and troughs in the period, following any recent, historic trends which develop.
A weighted average forecast can be calculated using a formula in Excel or by using a MRP system to perform the calculation.
| Month | Usage | Weighted Forecast |
|---|---|---|
| January | 450 | 450 |
| February | 190 | 450 |
| March | 600 | 294 |
| April | 600 | 478 |
| May | 420 | 551 |
| June | 380 | 472 |
| July Forecast | 440 | 417 |
Table 3: Weighted Average/Exponential Smoothing Forecasting