Good afternoon,
Whilst we accept that the statistical calculations housed within WFM are not readily available, we have some questions that we really need a better understanding on because the forecasting directly impacts our business.
We have imported historical data and now we are looking at forecasting for the month of August. Our historical data for the most recent period looks like this (for the activity in question):
12/06/2017 6
13/06/2017 9
14/06/2017 7
15/06/2017 7
16/06/2017 5
17/06/2017 3
18/06/2017 4
19/06/2017 2
20/06/2017 1
21/06/2017 2
22/06/2017 3
23/06/2017 7
24/06/2017 2
27/06/2017 1
28/06/2017 2
30/06/2017 2
01/07/2017 5
02/07/2017 2
03/07/2017 1
04/07/2017 2
05/07/2017 3
06/07/2017 2
07/07/2017 1
08/07/2017 4
09/07/2017 1
10/07/2017 7
11/07/2017 7
12/07/2017 8
13/07/2017 3
14/07/2017 1
15/07/2017 3
16/07/2017 3
17/07/2017 1
18/07/2017 3
20/07/2017 2
21/07/2017 3
23/07/2017 1
24/07/2017 4
25/07/2017 2
26/07/2017 1
If I look back over the past year there are only a handful of days when the daily IV was above 30 on any given day.
Yet the forecasting (using both UM and EA) is showing roughly the following for every week we have forecasted.
31/07/2017 - 37
01/08/2017 - 50
02/08/2017 - 43
03/08/2017 - 31
04/08/2017 - 27
05/08/2017 - 20
06/08/2017 - 13
This is using Universal Modeling and we have 14 months worth of data imported so there are no issues statistically.
Why is the system forecasting so high when we never see such high numbers? No overlays etc have been used, it's literally basing it solely on historical IV.
Any light you can shed would be greatly appreciated.
