Im using Trend Analysis - Double Exponential Smoothing Plot (not seasonal and it has a trend testedly) to forecast the net amount of carrier switchers (using Mobile number portability) in the future of a specific telecommunication provider. The data can be both negative and positive.
Accuracy Measures
MAPE 286.7 MAD 161.2 MSD 35891.4
All of the above measures are more than 100%. Am I using the wrong model?
MAPE is defined as
$$ \frac{1}{n} \sum_{i=1}^n \left|\frac{y_i - \hat{y_i}}{y_i}\right|\times 100 $$
so MAPE >100% means that the errors are "much greater" then the actual values (e.g. actual is 1, you predict 3, so MAPE is 200%). However beware that MAPE has many pitfalls as error measure, so often it won't be the best choice.
The question if you're using the wrong model or not, cannot be answered based on MAPE alone. Like any other error measure, there is no objective meaning of the values, you always interpret them relatively to some benchmark, e.g. MAPE of other model.
