As further support for mortgage scores being especially sensitive to number of accounts reporting balances:
1. My EX FICO 2 score didn't budge when 1 acct went from 0 to $10k, causing overall utilization to increase from 6% rounded to 9% rounded. Number of accounts reporting balances had not changed because another acct had dropped to 0 the same day. FICO 8, Auto 8, and Bankcard 8 all dropped significantly.
2. But then, a few days later, when 1 acct went from 0 to $4, causing number of accounts reporting balances to go up from 12/32 to 13/32, EX FICO 2 dropped 11 points. So for a crummy $4 I lost 11 points 
3. FICO 8, Auto 8, and Bankcard 8, of course, didn't react at all to that non-event.
Among other things this demonstrates the wisdom of implementing AZEO in the runup to a mortgage application.
Thank you for sharing this insight! Very helpful in figuring out methods to improve my mortgage scores.
I agree. They seem sensitive to CCs. I just lost 10 pts when 2 CCs ($500 & $5) reported a balance plus my LOC ($1). UTL well below 5% on all. Prior to that I had 1 CC and my LOC report (and my 2 loans). I have 8 CCs, 1 RC, 1 LOC, 1 mtg and 1 auto loan. This has happened before but just happened again.
For further confirmation that the difference between 12 accounts reporting balances and 13 accounts was worth 11 points on FICO 2 in my profile:
Today, when one 20% cccount dropped to zero, bringing the number of accounts with balances back down to 12, EX FICO 2 gave me back exactly 11 points, the same amount I lost when the number had gone up to 13.
In between 12/32 and 13/32 lurks 40%. So the question becomes, is 40% a thing, in EX FICO 2, when talking about number of accounts with balances?
Certainly looks like a solid datapoint.
For completeness how many of those 32 accounts are:
1) Open vs Closed
2) Revolvers vs Installment
Then which of the 12-13 balances were revolvers vs. installment?
Once I get through my current credit stuff I'll see about trying to nail down some breakpoints too on EX FICO 2, will state though EX FICO 2 is likely different than the FICO 04 varieties in this regard, EQ FICO 5 and TU FICO 4 both fall into that and I suspect the optimization goal is still the same: fewer accounts with balances = better.
That's sort of the end of the story from an optimization perspective haha.
Actually let me add a datapoint too: Number of Accounts with Balances in position 3 on my reason codes at a 760, though my file is a little strange cause God only knows how PLOC / HELOC are truly being counted after yesterday's datapoint but I'm fairly comfortable in suggesting they probably count as an account with balance at least. Also I do think installment are counted for this based on TT's and other's data I just haven't concretely confirmed that on my file though it certainly looks to be the case.
19 open accounts:
2 installment
1 HELOC
1 PLOC
14 Credit Cards
1 Amex (counts as a CC for EX FICO 2)
Balances on:
2 Installments
1 HELOC
1 PLOC
1 CC
5/19 of total accounts = 26.3%
3/17 of revolvers = 17.6%
Pretty sure 1/15 of CC's isn't an issue just sayin.
I will take this down to 2 installment, 1 HELOC, 1 CC and then finally 2 installment 1 CC and see if I can clear that reason code on EX FICO 2. Credit project for the next few months.
Also I'm adding on a pair of installment lines both from playing reindeer games in trying to find a new SSL trick provider, and I'll leave some small balance on them for at least a few months (First Tech sorta failed could only push it out 3 months no matter what I tried in their interface), that presumably will allow me to nail down installment vs. revolving on my file for this reason code explicitly.
@Revelate wrote:Maybe.
For completeness how many of those 32 accounts are:
1) Open vs Closed
2) Revolvers vs Installment
Then which of the 12-13 balances were revolvers vs. installment?
Once I get through my current credit stuff I'll see about trying to nail down some breakpoints too on EX FICO 2, will state though EX FICO 2 is likely different than the FICO 04 varieties in this regard, EQ FICO 5 and TU FICO 4 both fall into that and I suspect the optimization goal is still the same: fewer accounts with balances = better.
That's sort of the end of the story from an optimization perspective haha.
1. The 32 accounts are open. I haven't been referring to closed accounts at all.
2. Of the 32 one is an installment account (my "reindeer games" account), the other 31 are revolvers.
3. Of the 12 accounts with balances, one is the installment acount, the other 11 are revolvers.
4. If closed accounts are supposed to be factored into all of this, which would surprise me, my EX report has 27 zero balance closed accounts, 5 of which were installment accounts, and 22 of which were revolving accounts.
Right on bud, yeah I don't honestly think closed accounts were involved, just was a little surprised at the number of your accounts heh. Knew you had a busy file but that's really busy haha.
So looking at revolvers only:
11/31 ~35%
12/31 ~37%
I'm more inclined to think 40% breakpoint honestly than either of those, every breakpoint I found in FICO 8 for my file was on nice easy fractions: 1/3, 1/2 explicitly... 2/5 is an easy one too. Certainly quacks like a good datapoint, thank you for posting it!
I have the same outcome as SJ but with significantly less accounts. When I accidently had 5 of 12 report that's where I saw the dip. This includes 2 installments (home & auto) 2 CC and 1 LOC. Prior to this it was 4 of 12 reporting a balance, only 1 CC instead of 2.
@Revelate wrote:Right on bud, yeah I don't honestly think closed accounts were involved, just was a little surprised at the number of your accounts heh. Knew you had a busy file but that's really busy haha.
So looking at revolvers only:
11/31 ~35%
12/31 ~37%
I'm more inclined to think 40% breakpoint honestly than either of those, every breakpoint I found in FICO 8 for my file was on nice easy fractions: 1/3, 1/2 explicitly... 2/5 is an easy one too. Certainly quacks like a good datapoint, thank you for posting it!
Thanks.
I was initially talking about revolvers only, but @Anonymous advised me that installment accounts count in the metric of "accounts with balances" so I revised my focus.
At this point I may actually be oscillating between 12 and 13 for awhile, so I may really be able to pin those 11 points down. But I must confess that I would rather have only 2 or 3 accounts with balances, and leave the 40% experiment to others
Yeah it's busy all right. We're not even mentioning the 8 non-reporting, business accounts. But do keep in mind that I don't really have 32 reporting accounts, I only have 31. EX keeps thinking I have another open account which has been closed for awhile.