How are tracking signals typically computed?

The computation of tracking signals is primarily used to identify how well a forecasting method is performing in relation to the actual outcomes. The correct approach involves calculating the cumulative sum of forecast errors, which are the differences between the actual values and the forecasted values, and then this sum is divided by the Mean Absolute Deviation (MAD).

The rationale behind using the cumulative sum of errors is that it provides insights into systematic forecast bias—positive or negative. When this sum is divided by the MAD, it offers a standardized way to assess the magnitude of the errors in relation to the variability of the forecasted values. A tracking signal that falls outside predetermined control limits typically indicates that the forecasting process may need recalibration as it may not be performing adequately.

This method allows for continuous monitoring of the forecasting accuracy, making it a valuable tool in master planning of resources, particularly in supply chain and inventory management contexts. It helps organizations ensure that their demand plans are aligned with actual market conditions, allowing for necessary adjustments to be made timely and effectively.