AZEO vs 8% and 28%

[Anonymous]

Here you go read this and you’ll learn all kinds of stuff about it and everything else:

https://ficoforums.myfico.com/t5/Understanding-FICO-Scoring/General-Scoring-Primer-and-Version-8-Mas...

[Anonymous]

---

@Anonymous wrote:

  you want to practice AZEO with the one card having a balance under $100, so that you receive awards for optimizing: 1. utilization metrics, 2. raw dollar amount metric and 3. The number of accounts/revolvers with a balance metric.

---

Why does it need to be under $100? I thought it just needed to be under 9% (or some say under 3.99%). Is it proven that there's a raw dollar amount metric? I thought it had been proven that there wasn't...if there were, wouldn't extremely wealthy people be disadvantaged at all stages of utilization, as their dollar amounts for the respective percentage brackets would be much higher?

I've been letting a balance of ~$500 report on a card with a $30k limit, so that's about 2%. Are you saying if do less than $100 I'll get a higher score?

[Anonymous]

---

@Anonymous wrote:

---

@Anonymous wrote:

  you want to practice AZEO with the one card having a balance under $100, so that you receive awards for optimizing: 1. utilization metrics, 2. raw dollar amount metric and 3. The number of accounts/revolvers with a balance metric.

---

Why does it need to be under $100? I thought it just needed to be under 9% (or some say under 3.99%). Is it proven that there's a raw dollar amount metric?

I don't think this has been proven with FICO 8 Classic.

There are data points on small balance changes that affect the mortgage scores (EQ 5, TU 4, EX 2), like this post from HeavenOhio: Effect of low balances on my scores 

The ≤ 4% threshold was my own finding with a young credit history profile.  EX has a 4% aggregate utilization threshold (0,4] (EQ and TU also.)

It's a very small gain with FICO 8 Classic - EX +3, EQ +3, TU +1. The EX 8 line is a weird one though - all cards have to be at or below 4%, 4% aggregate alone won't work as it does with EQ/TU.

[Anonymous]

---

@Anonymous wrote:

---

@Anonymous wrote:

  you want to practice AZEO with the one card having a balance under $100, so that you receive awards for optimizing: 1. utilization metrics, 2. raw dollar amount metric and 3. The number of accounts/revolvers with a balance metric.

---

Why does it need to be under $100? I thought it just needed to be under 9% (or some say under 3.99%). Is it proven that there's a raw dollar amount metric? I thought it had been proven that there wasn't...if there were, wouldn't extremely wealthy people be disadvantaged at all stages of utilization, as there dollar amounts for the respective percentage brackets would be much higher?

I've been letting a balance of ~$500 report on a card with a $30k limit, so that's about 2%. Are you saying if do less than $100 I'll get a higher score?

---

@Anonymous Most people suggest a balance smaller than $100, even under $20 or even $5. We know there is a raw dollar metric on some versions and it may or may not exist on score 8. Nevertheless to ensure that you optimize that would be my advice, until further testing gives us a better understanding.

As to whether it would affect wealthy people more, first off it's a small penalty I think and no matter your income bracket more debt is more risk. So it makes sense to have small penalties as total debt increases.

@HeavenOhio  as well as @Anonymous  have proven this metric exists on some versions. Until we know more, test your profile and see. Matter of fact if you don't mind, come back and give us the data points.

[Anonymous]

And we don’t know where the thresholds are this is still in the beginning stages of study.

[K-in-Boston]

To clear up the earlier confusion about 8.9% vs 9.0%, the reason 8.9% is usually suggested is simply for a margin of error.  9.0% is fine, but 9.000001% rounds up to 10% on most models and doesn't get credit for being under 10%.

As for a raw dollar amount penalty, I don't believe there has been any evidence of that.  In the eyes of scoring models, if I have a $90,000 balance on $1.1 MM or so in revolving accounts, I'm far less of a default risk than the person who has a $50 balance on their sole $100 credit card.  There are of course issues with dollar amounts on charge cards on older models, where your balance isn't factored into utilization on say FICO 8 and 9, but whatever is in your "high balance" column is used as a credit limit on older models so someone who charges exactly $4000 a month on Amex Gold is always "maxed out."  (And the full balance is used for DTI calculations.)

[Anonymous]

Both @HeavenOhio and @Anonymous have evidence of it on certain versions. When I get a chance, I’ll find the links and post it for you @K-in-Boston .

[Anonymous]

---

@K-in-Boston wrote:

To clear up the earlier confusion about 8.9% vs 9.0%, the reason 8.9% is usually suggested is simply for a margin of error.  9.0% is fine, but 9.000001% rounds up to 10% on most models and doesn't get credit for being under 10%.

Is there proof of this happening? I have data points providing evidence for the use of a ceiling function. Testing greater than 8.99% but less than 9.00% 

I think this started based on a misunderstanding of the ceiling function, possibly combined with something about CPU floating point errors, which won't be an issue here.

[EDIT TO ADD: There won't be an inexact floating point result on the division because no one uses straight FP types when dealing with currency values. A decimal type would be used, or a method that traps the inexact error and corrects for it.]

Aggregate Util% = ( ( Balances / Limits ) * 100 ). If there is a fractional part, zero it and add +1. Ex: 5.16 -> 5.00 -> +1 = 6.00

There will always be an integer result. FICO's Score Estimator (not simulator) uses charts that suggest they only look at integers.

As for a raw dollar amount penalty, I don't believe there has been any evidence of that. 

Those reasons like 'High Revolving Balances' and 'Heavy use of revolving credit' only seem to show up on the older 5,4 and 2 models. Using AZEO for mortgage scoring purposes, it's not bad advice to suggest a low dollar amount.

I would think that would effect people with a long credit history a lot less than someone like me, with under 3yrs credit history.

[Anonymous]

@K-in-Boston

https://ficoforums.myfico.com/t5/Understanding-FICO-Scoring/Effect-of-low-balances-on-my-scores/m-p/...

@Anonymous could you link your findings regarding dollar amount thresholds please?

[Anonymous]

---

@Anonymous wrote:

@K-in-Boston

https://ficoforums.myfico.com/t5/Understanding-FICO-Scoring/Effect-of-low-balances-on-my-scores/m-p/5858091#M164008

@Anonymous could you link your findings regarding dollar amount thresholds please?

---

I didn't test it that throughly. It was just an observation about EQ 5 shifting reasons around below $1000 aggregate balance. Specifically, "High revolving balances" moved down a notch with a balance change, and aggregate 9% to 5%.