Regrettably, by the time TU finally reports my installment utilization drop from 89% to 8.9%, it will be impossible to get a clear read on the impact of the drop on the non-FICO 8 scores, due to other changes in the profile (account opening, balance changes, inquiries, etc). It's a shame, because it was a nice piece of intelligence for me to learn that in my profile FICO EX 2 is indifferent to installment utilization percentage, while FICO 8 found the change to be worth 31 points.
I'll report the score changes anyway, but they won't be ironclad, and I'll have to figure out and list what other factors may have been in play.
OK the results are in on my installment loan utilization dropping from 89% to 8.9% in TU report and the results are not identical to what happened in EX.
On FICO 8, as anticipated, the same 37 points I'd lost when the installment loan reported were regained.
With the other scores here's what happened or didn't happen:
FICO 4 +13
Auto 8 +58
Auto 4 +24
Bankcard 8 +56
Bankcard 4 +13
FICO 9 +36
Auto 9 +37
Bankcard 9 +34
So it seems clear to me that:
(1) my TU scores are more affected by installment utilization than my EX scores
(2) even my mortgage score jumped 13 points, while my EX score never budged
By comparison here was the reaction of the EX scores:
FICO 2 -- No change at all
Auto 8 -- +46
Auto 2 -- No change at all
Bankcard 8 -- +50
FICO 3 -- +11
Bankcard 2 -- -3
FICO 9 -- +27
Auto 9 -- +28
Bankcard 9 -- +24
FICO 8 -- +31
Although the TU results did not occur in the sterile environment of the EX test (report pulled 24 hrs after prior report with no other changes occurring), the only substantive change in TU was the addition of a new credit card account.
Just for the heck of it I decided to take the installment loan utilization percentage down from 8.9% to 4.8%, to see if there was some kind of 5% breakpoint.
And the answer is, there might be.
I was scratching my head over an unexplained 14-point EX FICO 8 gain, at a time when it seemed that only negative things were going on (new inquiry, new account, increased credit card utilization percentage), and noticed that the decrease in installment utilization had just been reported in EX.
Unfortunately, as Revelate has pointed out, my file is way too busy ordinarily to tell whether this was or was not a scoring factor, but it would be nice if someone with an SSL could test it in a more sterile environment one of these days.