@Anonymous wrote:
@GirlMelanie89 wrote:
@Anonymous wrote:
@GirlMelanie89 wrote:OP - I am wondering the same thing. As of now, my mortgage score is lower than my regular FICO score, but I just paid down my CC util from 83% to 1%. I am hoping to see an increase in my mortage score, because although my DTI is good, I cannot get approved if my mortgage score is not where it needs to be.
Are you referencing all of your scores or just one report? Just curious
Hi LuvTLH! Are you talking about in my siggy?
Yes. You don't mention Transunion. Are they your Fico8 scores or mortgage?
Ooohhhh. I guess I could have listed TU. I would get that score from my Walmart or Discover card. As of 2.12.16, my FICO 8 from them is 614. That should go up after my util reports and my late payment removal showing. As of 2.15.16, my mortgage score was 581. 
I usually dont list my TU score, because I cant update it daily as I can with EQ and how it updates when anything on my report changes.
@iv wrote:
@Anonymous wrote:
Marketing fluff or not.
Do you have any evidence backing this claim, or are you just making an assumption as well? I can't think of any. I think we can agree that lower utilization is not one of them.Sure.
It's even fairly straight-forward.
There are certainly some broad generalizations you can make about most models (adding new derogatory account markers to a previously clean file is going to be negative; using 100% of open credit lines is going to be negative...)
But many "single" actions you can take with your credit result in mutliple changes to the report, some with generally postive effects, and some with generally negative effects. Given that different models, industry options, and scorecards weight those effects differently (even if they all consider the same set as postive vs negative), it's easy to see how not only the magnitude, but also the direction of the score change can differ between models.
(And no, I can't agree that lower utilization will always increase score on all models.)
A few simple examples of common actions that can result in some score models moving in opposite directions:
You are carrying revolving debt at a 50% utilization ratio, and then open a new card that doubles your total credit lines.
Result: New inquiry (negative or neutral); New account (negative or neutral); reduced AAoA (negative or neutral); reduced utilization to 25% (postive or neutral)
You have a old charged-off account (originally opened 16 years ago), and only one other account on your report (opened this year, postive history), and then the charged-off account is removed from the report.
Result: Clean report (postive, but also changed scorecard); reduced AAoA from 8 years to 1 year (negative); reduced age of oldest account from 16 years to 1 year (negative)
You are carrying revolving debt at a 5% utilization ratio, and then pay it all off, allowing 0% to report on all cards
Result: Lower utilization (slight positive or neutral at this level); Showing no use of revolving credit (negative)
You are carrying revolving debt at a 20% utilization ratio, and then pay it all off, allowing 0% to report on all cards
Result: Lower utilization (positive); Showing no use of revolving credit (negative)
Notice that in all these cases, there are both postive and negative pressures on the score from a single consumer action.
It is quite easy for different models (even those with exactly the same idea of what a "positive" change or a "negative" change is), to weight those changes differently.
A model that weights the negative factors in a particular scenario higher, and the positive factors lower may result in a score decrease; while a model that weights the same positive factors higher, and the negative lower may result in a score increase.
Not to mention that some models do consider certain things negative/postive that other models just disregard, and that different models have scorecard breakpoints at different criteria.
Among the mortgage tri-merge scores, EXv2 is most notable for moving differently than FICO 8, or EQ5/TU4 for the same report changes.
What'd I miss? FICO 98 (EX FICO Risk Model v2 base) does count installment utilization like FICO 8, whereas the FICO 04 variants don't. I don't play many games with my revolving utilization if we're referring to that, didn't get useful datapoints on that during my own mortgage process for admittedly good reason.
Generally speaking though, I'm comfortable with saying with regards to FICO algorithms: less but non-zero revolving utilization is "better" on any version of the FICO algorithm we're familiar with; that said, I absolutely agree that one can't take movements in one algorithm to be 1:1 with any other algorithm, but in broad strokes this holds true across a wide swath of anecdotal evidence. Sure 20% -> 1% might not gain you anything depending on algorithm, but 83% to 1% is a pretty large swing assuming other things are held constant regardless of individual model.
It's much like saying fewer inquiries are better, even if I don't think on EQ Beacon 5.0 there's any difference between 0 inquiries and 2 inquiries on my file, and admittedly I can't generalize that to anyone else's file either as I am stuck in the dirty buckets.
And if we want to get really goofy when talking score versions: all my other scores are pretty consistent across the board on a 3B report (with some variations due to individual algorithm): my EQ FICO 9 is a 780, which is VS 3.0 territory; however, my EX FICO 9, is right around a 700 in line with EX FICO 8 (and EQ FICO 8 too) with nearly identical information other than inquiry difference, most notably an inquiry a few days ago on EX. So yeah, we can't take any given score as a predictor of any other score, and absolutely the breakpoints and weightings are different, but we can make some general assumptions which have held true throughout all the anecdotal data I've seen at least.
