@Thomas_Thumb wrote:I suspect the key thing for stability (on Fico 8 and Fico 9) is no new accounts. My last new account was 2011. Of course, aggregate UT% needs to be kept low regardless of scorecard/profile. My current AAoA is approximately 19 years with AoOA at 35 years. As mentioned in other threads, I see score swings in the older Classic Fico 04 and Fico 98 models depending on # cards reporting. The swings are rather pronounced on EQ Fico 04. Open mortgage loan was above 50% at the beginning of 2014. Now it's at 20%.
..
I'm definitely of the opinion that on FICO 8/9 new accounts are far more impactful than rationally expected or historically noted: not certain it's confined to just AOYA yet but maybe, don't think I ever had an account report as quickly as Citi did while on a clean scorecard and as such might not have seen such a wild swing as before given there may be breakpoints sub 1 year.
Congrats on the mortgage paydown progress, definite win there; also which service did you pull that nice format by age for accounts? I'm still tracking it all on a spreadsheet but it'd be nice to occasionally get a prettier view that presumably I can click on for further tradeline details.
While it would be ideal to see who has the lowest AoOA at exactly 850, perhaps another approach would be to find who has 840+ and low AoOA. Finding those profiles, understanding the relative optimization, can give insight to whether AoOA is acting as the limiter.
As those members report their ability to further increase score, or if they get only small points increases while otherwise optimized, at say 13 years AoOA, that data can be just as useful.
Arrgh! So close, and yet, so far!
From USAA as of May 13, 2019. It's torturing me! Naah, not really. A humble brag, I know, but I've worked damned hard to get my score where it now is. My oldest account would have to be American Express from 1987.
@Anonymous wrote:
I think that’s a really great idea NRB525. Maybe I should start a separate thread for that? Collect the 840s there while BBS is collecting 850s.
Well if we're going to do that why not extend it even further?
I still submit my analysis that when if I had zero scorable inquiries and no new accounts I'd be non-trivially higher than my current 794 EX score, probably to my record 827 and actually beyond soon north of 830 with the AAOA 5 year mark.
Then the question really becomes how much my CFA counts for, and how much I'm losing on installment utillization as a result of my mortgage. I'm nearly 100% confident that it counts somehow, but we'll see as I continue to pay down my auto loan but I know I got 20+ points when I had my tax lien, if it works the same in my scorecard once the new accounts be gone, that sounds like an 850 frankly or otherwise hitting some scorecard cap if AOOA really is a major factor on FICO 8/9.
@Anonymous wrote:
Sounds great and I totally agree with everything you said. How do you suggest we widen it? what parameters do you suggest?
This is all back of hand, it bears more discussion I suspect; that said, I think there needs to be a floor around 820, maybe 810.
New file scorecards (call it sub 3 years) seem to cap out around 800ish after 2 years, not sure we've had anyone establish say 4 revolvers and one SSL trick loan immediately and then sit on their hands for 3 years to get a good benchmark there. I'm also making the blithe assumption that a 30D late scorecard can't get to 820 either.
We'd need the 3 aging metrics:
AOYA (and admittedly this might be pattern under FICO 8, still think we're unsure of that)
AAOA
AOOA
Open installment loans / aggregate utilization I think is a useful metric.
This also presumes revolvers optimized AZEO style for an optimal datapoint, which suggests total number of open revolvers might be useful datapoint wise too.
@Anonymous wrote:
Again I totally agree and it sounds good I also think it merits more thought and discussion as you said. Do we limit it to people under a certain AooA? Do we try to find someone relatively new that’s willing to do the test maybe a few people?
Nah open it up to everyone, really it's just a data collection and analysis exercise.
You might want to search on ABCD's name he did some analysis I don't think quite along these lines but potentially.