Any suggestion how to reduce the Nested Loop performance?
The 9,788 rows read from the Index Seek are replayed by the spool once for each row arriving at the Nested Loops join's top input. 9788 * 13,070 = 127,929,160.
The Nested Loops join applies the following predicate on each replay:
A better strategy would be to perform a correlated seek, with no join residual. You should check the indexes available on the table underlying this view to make sure the correlated seek strategy is viable.