topic Re: Utilization DP in Understanding FICO® Scoring

Utilization DP

keekers — Sat, 27 Jul 2024 23:39:55 GMT

FWIW, my utilization increased from 4% to 5% and I lost 2 points on Experian. It will be dropping back to at least 4% in the next week or so; I'm curious if I'll get those points back once my next card reports for this month. To be continued.

Re: Utilization DP

MissLiz — Sat, 27 Jul 2024 23:01:02 GMT

You didn't see a score change on TU or EQ as well?

Re: Utilization DP

keekers — Sat, 27 Jul 2024 23:04:06 GMT

@MissLiz wrote:
You didn't see a score change on TU or EQ as well?

I only use free services to track my FICO scores; EQ and TU haven't updated quite yet. Will report once they do.

Re: Utilization DP

JoeRockhead — Sun, 28 Jul 2024 01:17:01 GMT

5% shouldn't be a threshold but, all the major CRAs have their own tweaks to the scoring models. I was rewarded with a 1 point drop on TU Fico 8 for getting a CLI.... ¯\_(ツ)_/¯

You'll get your points back.

Re: Utilization DP

SouthJamaica — Sun, 28 Jul 2024 11:42:06 GMT

@keekers wrote:
FWIW, my utilization increased from 4% to 5% and I lost 2 points on Experian. It will be dropping back to at least 4% in the next week or so; I'm curious if I'll get those points back once my next card reports for this month. To be continued.

1. 2 points is not statistically significant.

2. Your EX scores don't fluctuate that often based on slight changes in aggregate utilization.

So something else caused it. But it might be undetectable.

Re: Utilization DP

keekers — Thu, 08 Aug 2024 21:36:59 GMT

Update: I understand that 5% may not be a threshold but my EQ score dropped 5 points when my utilization went up to 5% and as of yesterday I got the 5 points back when it dropped back to 4%. Nothing else changed that I can tell. EX still shows the loss of 2 points and still shows 5% UTI since they aren't reflecting the July balance decrease for some reason. TU didn't budge with either movement up or back down. *SHRUGS*

Re: Utilization DP

SouthJamaica — Fri, 09 Aug 2024 03:21:03 GMT

@keekers wrote:
Update: I understand that 5% may not be a threshold but my EQ score dropped 5 points when my utilization went up to 5% and as of yesterday I got the 5 points back when it dropped back to 4%. Nothing else changed that I can tell. EX still shows the loss of 2 points and still shows 5% UTI since they aren't reflecting the July balance decrease for some reason. TU didn't budge with either movement up or back down. *SHRUGS*

In order to be able to draw such conclusions you would need to know

-the date the score changed

-the date the reported utilization changed

-the overall data from your report the day before the score changed

-the overall data from your report the of the score change

I'm pretty sure you know none of the above.

Re: Utilization DP

Patient957 — Fri, 09 Aug 2024 15:46:18 GMT

@SouthJamaica wrote:
1. 2 points is not statistically significant.

I think it's worth pointing out that statistical significance has nothing to do with this. There's nothing random about FICO scoring models -- they are deterministic.

Re: Utilization DP

Thomas_Thumb — Fri, 09 Aug 2024 19:29:38 GMT

Uncontrolled variables are important considerations. This includes # of cards reporting balances, utilization levels on each card with a balance, total amount owed, scorecard assignment (this is a big one for the generalists that think what happens to me will happen to everyone) and aging factors.

You can't have an increase in aggregate utilization without an increase in total $ owed. You can't have an increase without some card showing an increased balance. Is the card with the increased balance also the one with highest utilization?

Why aren't Fico scores on all CRAs showing a score change? I do know # cards with balances can be tweaked as a factor by CRAs and impact is different among the various Fico models. Moreso on older Fico "mortgage" models. Allowed tweaking was curtailed on Fico 8 and further restricted with Fico 9.

What defines a threshold? A significant change in outcome - imo. What is significant in Fico scoring. I'd say a change of 5 or more points might qualify. If the event does not have a repeatable effect or the effect is not observed across all CRAs (showing the event) then, I wouldn't call it a general threshold.

A youngest revolving account crossing above/below 12 months is a general threshold.

Re: Utilization DP

keekers — Sat, 10 Aug 2024 15:45:11 GMT

Is there a thread or a central place where the rules of thumb, thresholds, etc. are located for reference?

Re: Utilization DP

Thomas_Thumb — Sun, 11 Aug 2024 06:26:32 GMT

@keekers wrote:
Is there a thread or a central place where the rules of thumb, thresholds, etc. are located for reference?

Nothing concise. There is a "credit scoring primer" thread 509 posts long that attempts to pull information together. It does have a lot of useful information. However, I'd hardly call that concise. Do a search and it will come up.

There is no hard definition on what constitutes a Fico score threshold. Kinda like no definition on what might be considered a pain threshold and its association to risk of severe injury.

A child whose finger is pricked and starts crying might indicate a finger puncture is a significant pain threshold. However, an adult might have a finger severed and not cry (likely yell and/or cuss instead). So, how does one identify pain thresholds? Should there be scorecards (young child vs adult) where response level depends on scorecard assignment? Should vocalized decibel output be the indicator for establishing pain thresholds? If yes, what Dba change would be considered a significant threshold for pain as it relates to risk of injury being severe?

Fortunately with credit models the output is numerical. New inquiries, if/when one impacts score, typically causes a 5-7 point drop. A change of that magnitude is sufficient for assessing inquiry count thresholds. So, a change in score of 5 or more points is often used for threshold analysis.

Just a note on inquiries: they can be deduped so a group of them count as one for scoring purposes. They can also be tagged with a 30 day buffer so no impact on score until 30 days has passed. Inquiries are also subject to binning such as : 0,1,2,3,4-5,6-9 and 10+. In this case # of scoreable inquiries needs to change bins to affect score.

Re: Utilization DP

AzCreditGuy — Sun, 11 Aug 2024 23:34:41 GMT

You want to test Utilization DP? I am moving around money and its interesting how pts go up and down

Re: Utilization DP

Thomas_Thumb — Mon, 12 Aug 2024 01:31:05 GMT

@AzCreditGuy

Nice datapoints.

EQ increased 24 points when aggregate UT dropped from 10.6% to 2.7%. Highest card UT was maintained at 51.7%. This isolates aggregate UT from highest card UT. The large increase in score suggests an AG utilization threshold may have been crossed (at 9%, 5% or both?)

Conversely, TU and EX scores both dropped 19 points although AG UT went from 10.6% to 10.4%. Why? Because highest card UT increased from 51.7% to 88.8%. This clearly isolates AG UT from highest individual card UT. The score drop shows how a change in highest individual card utilization can also impact score. The large score drop suggests a threshold (at 69%?) was crossed.

As a further test can you drop TU and EX aggregate UT below 9% while maintaining highest card at 88.8%? Would it be possible to do this by paying down the 51% UT card? Then check impact on score and provide an update.

Re: Utilization DP

AzCreditGuy — Wed, 14 Aug 2024 21:10:50 GMT

@Thomas_Thumb wrote:
@AzCreditGuy
Nice datapoints.

EQ increased 24 points when aggregate UT dropped from 10.6% to 2.7%. Highest card UT was maintained at 51.7%. This isolates aggregate UT from highest card UT. The large increase in score suggests an AG utilization threshold may have been crossed (at 9%, 5% or both?)

Conversely, TU and EX scores both dropped 19 points although AG UT went from 10.6% to 10.4%. Why? Because highest card UT increased from 51.7% to 88.8%. This clearly isolates AG UT from highest individual card UT. The score drop shows how a change in highest individual card utilization can also impact score. The large score drop suggests a threshold (at 69%?) was crossed.

As a further test can you drop TU and EX aggregate UT below 9% while maintaining highest card at 88.8%? Would it be possible to do this by paying down the 51% UT card? Then check impact on score and provide an update.

The Cap 1 is only near maxed at the moment, as they offer a 0% fee to Transfer money into accounts for a BT. I will never use it for a BT since the apr is 29%, I only use it to have that money in accounts for a month or so and pay it off before the free interest grace period is over. The Cap 1 will be back to 0% UTI after 08/28/24

From the prevoius DP points, it was only a matter of Experian catching up and reporting the balance on my Cap 1. It seems Experian is the worst drop off at 39 pts vs 19 from EQ and Trans.

Re: Utilization DP

AzCreditGuy — Fri, 16 Aug 2024 20:33:10 GMT

Heres another interesting DP in just 3 days my EX and EQ scores have went up 9 to 10 pts. What happened? I believe this is due to the UTI with Citizens going from 51.4% down to 49.86% What I find interesting is my Wells UTI went from 0% to 13.02% for UTI.....Maybe I could have gotten more pts up, if that wasnt there?

Re: Utilization DP

AzCreditGuy — Tue, 27 Aug 2024 16:20:23 GMT

Another interesting data py between EQ and TU....Seems TU rewards you for more debt vs EQ taking pts away for debt...EQ -3 pts TU 7+ pts ...Ex is not affected