RDF Promo functionality gives too high forecast

 Hello to all participants!I faced the following problem while testing of RDF promo functionality.

1.       Simple case.

a) we have constant regular sales during some periods (for previous 56 weeks) i.e RSAL=C.

b) There only promo was here the only time - two weeks ago and its effect=E (in pcs.)

c) The same promo is planned for the next week for same item in same store.

d) The forecast method is causal,  Promo Effect Type is set to “Automatic”.

What I expect to receive for 57^th week:

·         RSAL = C

·         Approved Promo Effect = E

·         Promoted forecast = C+E.

What I really receive: