cancel
Showing results for
Did you mean:

## Utilization Intervals, Brackets, etc.

Auto Loans for ANY Credit Situation. Immediate Response.
Established Contributor

## Utilization Intervals, Brackets, etc.

[Yes, I did more than due diligence using Google site search, and I'm still uncertain of a few minor things: site:ficoforums.myfico.com +utilization. I even used Google images site search for any image related to utilization posted on these forums. Plenty of results to wade through.]

Is this the best aggregate utilization percentage interval:  0 < AggrUtil < 9 , or in standard mathematical interval notation,  (0,9) ?

Or is it 0 < AggrUtil < 10 or (0,10)?

As of today, my score has been calculated based on a 0.47% real utilization using 40/8500 and 1 of 2 cards with a balance. So if my next statement balances were to show 179.90 and 584.68 for balances on my 2 cards with a combined credit limit of 8500, would I cross into the next [9,29) interval with this new 8.995% aggregate utilization, and experience a loss of points on FICO 8?

I've found the following information on the forums, all from Senior Contributors or higher:

1) Total utilization breakpoints: 9%, 29%, 49%, 69%, 89%

2) Individual utilization breakpoints: 29%, 49%, 69%, 89%

3) A whole lot of nitpicking about 1 and 2

4) There's breakpoints below 9% as well.

5) There's no breakpoints below 9% - something else must have caused it!

6) The breakpoint is actually at 10%.

7) There's a breakpoint at 5%.

8) Utilization division calculations are not rounded, only truncated.

9) Utilization division calculations are rounded to the next highest integer.

10) Utilization division calculations are rounded to the nearest integer.

11) No solid information on the type of rounding method.

12) "This comes up dozens of times a day and it's already well settled..."' (Sure, if you've been following the discussion incrementally for the last 2+ years, maybe it has been settled.)

And then I find information from FICO itself like the first 2 graphics below:

And here's 2 more charts that have been already been posted on these forums in the past:

[Credit: @Thomas_Thumb ] Reportedly based on a one card account where aggregate utilization = individual card utilization. This data was from a non Fico blog that a MyFico member posted here. @Thomas_Thumb plotted the data to create the graphical display, and previously posted it on the forums.

Source seems to be: Bob Wang

New CRADIS contacts, bearing 1 carom 10. Just think of EQ, TU, and EX as Cylon Basestars.

Message 1 of 13
12 REPLIES 12
Super Contributor

## Re: Utilization Intervals, Brackets, etc.

My comments below in blue.  Hope this helps.

@CassieCard wrote:

[Yes, I did more than due diligence using Google site search, and I'm still uncertain of a few minor things: site:ficoforums.myfico.com +utilization. I even used Google images site search for any image related to utilization posted on these forums. Plenty of results to wade through.]

Is this the best aggregate utilization percentage interval:  0 < AggrUtil < 9 , or in standard mathematical interval notation,  (0,9) ?

Or is it 0 < AggrUtil < 10 or (0,10)?

FICO models round all percentages up.  Therefore a util percentage of 0.23% (for example) is considered by FICO as 1%.  The best aggregate or total utilization is 1-9%.  Just bear in mind that a utilization of 9.01% is rounded up to 10%, at which point a penalty is levied.

As of today, my score has been calculated based on a 0.47% real utilization using 40/8500 and 1 of 2 cards with a balance. So if my next statement balances were to show 179.90 and 584.68 for balances on my 2 cards with a combined credit limit of 8500, would I cross into the next [9,29) interval with this new 8.995% aggregate utilization, and experience a loss of points on FICO 8?

Your total utilization would still be in the 1-9% range.  If it went up to 9.01%, it would be rounded up to 10%, and therefore a penalty would be levied.

I've found the following information on the forums, all from Senior Contributors or higher:

1) Total utilization breakpoints: 9%, 29%, 49%, 69%, 89%

Yes

2) Individual utilization breakpoints: 29%, 49%, 69%, 89%

Yes.  These are true.  Some evidence exists that a tiny penalty for indiviudal U might exist at 9% as well.

3) A whole lot of nitpicking about 1 and 2

4) There's breakpoints below 9% as well.

These may exist but I am skeptical.  If so, even the people who think they are likely to be real would tell you that they involve tiny amounts of points and often do not affect everyone.

Part of the reason I am skeptical is that we haven't had careful controlled tests (run by several people with different profiles) to exclude the possibility that some FICO models care about the actual raw dollar values of debt, which FICO's published language implies it does care about.  (Implies without explicitly and incontrovertibly stating so, thus the disagreement on whether this happens.)

If so this might explain why people with a low credit limit see no score change between a total U of 0.3% and 8.7%, whereas some people with a gigantic credit limit do seem to see a change.  The person with the gigantic limit is increasing his dollar value of debt by a huge amount whereas the person with the low limit is not.

5) There's no breakpoints below 9% - something else must have caused it!

I am skeptical, but it is possible.

6) The breakpoint is actually at 10%.

No.  This is wrong.

7) There's a breakpoint at 5%.

I am skeptical, but it is possible.

8) Utilization division calculations are not rounded, only truncated.

No

9) Utilization division calculations are rounded to the next highest integer.

Yes

10) Utilization division calculations are rounded to the nearest integer.

No

11) No solid information on the type of rounding method.

No

12) "This comes up dozens of times a day and it's already well settled..."' (Sure, if you've been following the discussion incrementally for the last 2+ years, maybe it has been settled.)

And then I find information from FICO itself like the first 2 graphics below:

And here's 2 more charts that have been already been posted on these forums in the past:

[Credit: @Thomas_Thumb ] Reportedly based on a one card account where aggregate utilization = individual card utilization. This data was from a non Fico blog that a MyFico member posted here. @Thomas_Thumb plotted the data to create the graphical display, and previously posted it on the forums.

Source seems to be: Bob Wang

One of the most important issues (which you do not mention) is that in practice these distinctions rarely matter.  Typically a person either needs some extra scoring points, due to an upcoming application, or he doesn't.  If he does, he can pay all cards to zero except one with the remaining card showing a small positive balance.  That will always optimize his score as touches the factors of CC balances.

And if he doesn't need any extra points in the near future, he can ignore the whole idea of ultra-low utilization.  Having an extremely low utilization every single month will not help him build a higher score any faster than if his utilization is occasionally 0%, 15%, 25%, or even 45%.  (It's a good idea to keep total U under 29% and individual U under 49%, just in case a creditor would be freaked out by something higher, but the chance of this happening is low -- and it is almost zero if we are talking about a one-month blip of something high.)

Message 2 of 13
Senior Contributor

## Re: Utilization Intervals, Brackets, etc.

Very intresting

All I know is I have been at 8%utilization and went to 1%utilization,  zero change. I always believe that threshold under 9%, but going lower doesn't gain a thing.

Message 3 of 13
Senior Contributor

## Re: Utilization Intervals, Brackets, etc.

@CassieCard wrote:

Is this the best aggregate utilization percentage interval:  0 < AggrUtil < 9 , or in standard mathematical interval notation,  (0,9) ?

Or is it 0 < AggrUtil < 10 or (0,10)?

As of today, my score has been calculated based on a 0.47% real utilization using 40/8500 and 1 of 2 cards with a balance. So if my next statement balances were to show 179.90 and 584.68 for balances on my 2 cards with a combined credit limit of 8500, would I cross into the next [9,29) interval with this new 8.995% aggregate utilization, and experience a loss of points on FICO 8?

I've found the following information on the forums, all from Senior Contributors or higher:

1) Total utilization breakpoints: 9%, 29%, 49%, 69%, 89%

2) Individual utilization breakpoints: 29%, 49%, 69%, 89%

3) A whole lot of nitpicking about 1 and 2

Fico considers three attributes that relate to revolving credit use as follows:

a) Aggregate utilization (all accounts combined)

b) Number and/or % of open accounts having a balance

c) Individual card utilization.

Aggregate utilization is weighted more heavily than individual card utilization or # accounts with balances on Fico 8. Number of accounts with balances has a relatively small effect on Fico 8. Unlike Fico 8, the older Fico "mortgage" models do put significant weight on # of accounts with balances - and presence of a recent inquiry may increase weighting further.

In your example above you are increasing #/% of revolving accounts with balances and this could cause a score drop of up to 5 points - IMO. Is your aggregate utilization crossing above the lowest threshold? - possibly. If so, that could ding score 5 points. Everything considered, your score could drop 5 to 10 points. Of course, this is only a point in time impact on score. So, no need to be concerned.

Note what happens, report back and record the result for future reference.

As a general rule I recommend managing credit to maintain aggregate utilization under 9% and individual card utilizations under 29%. That being said,  experiencing a spike in aggregate UT to the 9% - under 29% range and/or individual card utilization into the 29% - under 49% range is not a problem, particularly if balances are PIF. Unlike lates, there is no lingering impact on score.

FYI - The utilization grapg you pasted (with the step changes in score) was reportedly based on a one card account where aggregate utilization = individual card utilization. So, impact is due to affect of both scoring attributes. This data was from a non Fico blog that a MyFico member posted here. I plotted the data to create the graphical display.

Fico 8: .......EQ 850 TU 850 EX 850
Fico 9: .......EQ 850 TU 850 EX 850
Fico 4 .....:. EQ 809 TU 823 EX 830 EX Fico 98: 842
VS 3.0:...... EQ 835 TU 835 EX 835
Fico 8 BC:. EQ 892 TU 900 EX 900
CBIS: ........EQ LN Auto 940 EQ LN Home 870 TU Auto 902 TU Home 950
Message 4 of 13
Established Contributor

## Re: Utilization Intervals, Brackets, etc.

@CreditGuyInDixie wrote:

My comments below in blue.  Hope this helps.

@CassieCard wrote:

Is this the best aggregate utilization percentage interval:  0 < AggrUtil < 9 , or in standard mathematical interval notation,  (0,9) ?

Or is it 0 < AggrUtil < 10 or (0,10)?

FICO models round all percentages up.  Therefore a util percentage of 0.23% (for example) is considered by FICO as 1%.  The best aggregate or total utilization is 1-9%.  Just bear in mind that a utilization of 9.01% is rounded up to 10%, at which point a penalty is levied.

Now I see why there is confusion around that 10% breakpoint: there's a pre-rounding interval and post-rounding interval. The FICO module is essentially using the equivalent of the Excel Ceiling( (Sum of Statement Balances / Sum of Credit Limits ) * 100, 1) function to get an integer. (No, I am not suggesting FICO uses Excel. lol More people are familiar with Excel functions than something like std::ceil in C++, which is what a high performance FICO scoring cluster should be using.)

So the best value to have for ( (Sum of statement balances / Sum of credit limits ) * 100 ) lies in the (0, 9] interval. And now I can also see why you always recommend staying under 8.99%.

@CassieCard wrote:

As of today, my score has been calculated based on a 0.47% real utilization using 40/8500 and 1 of 2 cards with a balance. So if my next statement balances were to show 179.90 and 584.68 for balances on my 2 cards with a combined credit limit of 8500, would I cross into the next [9,29) interval with this new 8.995% aggregate utilization, and experience a loss of points on FICO 8?

Your total utilization would still be in the 1-9% range.  If it went up to 9.01%, it would be rounded up to 10%, and therefore a penalty would be levied.

So 8.995 is in the optimal (0,9] interval. 9.01 is in the next (9, 29] interval. I know it's not a lot of points we're talking about here, but so many of the basic 'how to improve your credit' articles say only to 'keep your utilization below 10%', which is really off by 1 for the most optimal outcome.

@CassieCard wrote:

I've found the following information on the forums, all from Senior Contributors or higher:

1) Total utilization breakpoints: 9%, 29%, 49%, 69%, 89%

Yes

Or 7 scoring variations: 0% and 6 intervals: (0,9]%, (9, 29]%, (29,49]%, (49, 69]%, (69, 89]%, (89, ∞)% (I'm intentionally not writing this out in set notation.)

This might explain some posts that state something like, "I dropped utilization below 30% and my score didn't change." Maybe it was actually 29.1%, which is in the same interval, and the presentation software rounded down to 29%. I certainly see it on my Experian CreditWorks report where 0.47% was rounded down to 0% in one section.

@CassieCard wrote:

7) There's a breakpoint at 5%.

I am skeptical, but it is possible.

If it exists on my thin file score card, I should find out in April. I'll go from 7/7/7 utilization on next week's (February) report to 6/6/6 in March and then 5/5/5 in April.

The only thing changing between those reports will be ageing, and the next threshold for that is 7 months away. (AAoA is currently 5 months.)

@CreditGuyInDixie wrote:

One of the most important issues (which you do not mention) is that in practice these distinctions rarely matter.

I'm aware of how ridiculous my year-long low-utilization experiment must seem to many. It's just an informal hypothesis test that even if shown to be correct doesn't prove anything. When it's over, I plan on blaming Reddit and forgetting it ever happened. I just needed more detail about the utilization calculation. I went from a fascination with the R programming language to a fascination with R and FICO scoring. It's all academic really. The next time I need to use my FICO score for anything might be in 2021 if I am lucky and nothing happens to my car.

@CreditGuyInDixie wrote:

Typically a person either needs some extra scoring points, due to an upcoming application, or he doesn't.  If he does, he can pay all cards to zero except one with the remaining card showing a small positive balance. That will always optimize his score as touches the factors of CC balances.

And if he doesn't need any extra points in the near future, he can ignore the whole idea of ultra-low utilization.  Having an extremely low utilization every single month will not help him build a higher score any faster than if his utilization is occasionally 0%, 15%, 25%, or even 45%.  (It's a good idea to keep total U under 29% and individual U under 49%, just in case a creditor would be freaked out by something higher, but the chance of this happening is low -- and it is almost zero if we are talking about a one-month blip of something high.)

I believe you. I'll definitely get another card at least a year before applying for a mortgage, and then use the AZEO method when it's needed. But for now, anything above 8% of my total credit limt is as useful as Monopoly money to me.

Thanks a lot for your always helpful information! I just wanted to bring as much utilization info together somewhere so that someone else new to credit might find it all more easily.

New CRADIS contacts, bearing 1 carom 10. Just think of EQ, TU, and EX as Cylon Basestars.

Message 5 of 13
Established Contributor

## Re: Utilization Intervals, Brackets, etc.

@Thomas_Thumb wrote:

@CassieCard wrote:

As of today, my score has been calculated based on a 0.47% real utilization using 40/8500 and 1 of 2 cards with a balance. So if my next statement balances were to show 179.90 and 584.68 for balances on my 2 cards with a combined credit limit of 8500, would I cross into the next [9,29) interval with this new 8.995% aggregate utilization, and experience a loss of points on FICO 8?

I've found the following information on the forums, all from Senior Contributors or higher:

1) Total utilization breakpoints: 9%, 29%, 49%, 69%, 89%

2) Individual utilization breakpoints: 29%, 49%, 69%, 89%

3) A whole lot of nitpicking about 1 and 2

Fico considers three attributes that relate to revolving credit use as follows:

a) Aggregate utilization (all accounts combined)

b) Number and/or % of open accounts having a balance

c) Individual card utilization.

Aggregate utilization is weighted more heavily than individual card utilization or # accounts with balances on Fico 8. Number of accounts with balances has a relatively small effect on Fico 8. Unlike Fico 8, the older Fico "mortgage" models do put significant weight on # of accounts with balances - and presence of a recent inquiry may increase weighting further.

Does 'older FICO mortgage models' include EX FICO 2? That's my second highest score right now at 727, with [Short credit history, Short revolving history, Too many accounts with balances, Recently opened too many new credit accounts] for score factors.

@Thomas_Thumb wrote:

In your example above you are increasing #/% of revolving accounts with balances and this could cause a score drop of up to 5 points - IMO. Is your aggregate utilization crossing above the lowest threshold? - possibly. If so, that could ding score 5 points. Everything considered, your score could drop 5 to 10 points. Of course, this is only a point in time impact on score. So, no need to be concerned.

Note what happens, report back and record the result for future reference.

Each month a new myFICO 3B report will be posted as a reply in HTML table format to my 'All at Just Under 8.99% Uitilization Experiment' post. There's already changes to a lot of scores there starting from December 4th with 1 open loan at 17% util through a closed loan and my first 2 credit cards ever being added with a combined \$8500CL. This next 2% on 1 with 0.47% aggregate to 7% on 2 with 7% aggregate change will start to happen next week, with Experian CreditWorks Premium updating first. I have to wait for TU to update (took 6 days extra last time) before I generate my monthly 3B+ report on or around Feb 11.

Based on everything I've read from you and others, my score shouldn't change very much over the next 3 months with steadily decreasing utilization from 7/7/7 to 6/6/6 and finally 5/5/5. But this is also on a score card where 1 inq causes a 19pt drop and adding the first revolving account ever causes a +51pt gain, so who knows.

@Thomas_Thumb wrote:

As a general rule I recommend managing credit to maintain aggregate utilization under 9% and individual card utilizations under 29%. That being said,  experiencing a spike in aggregate UT to the 9% - under 29% range and/or individual card utilization into the 29% - under 49% range is not a problem, particularly if balances are PIF. Unlike lates, there is no lingering impact on score.

A year from now I will be well prepared to start using my cards like a normal human due to people like you on these forums. I would have always paid in full each month, but I wouldn't have known how to optimize utilization, and a whole bunch of other things related to mortgage applications.

@Thomas_Thumb wrote:

FYI - The utilization graph you pasted (with the step changes in score) was reportedly based on a one card account where aggregate utilization = individual card utilization. So, impact is due to affect of both scoring attributes. This data was from a non Fico blog that a MyFico member posted here. I plotted the data to create the graphical display.

I will credit you and list that information in the image caption!

New CRADIS contacts, bearing 1 carom 10. Just think of EQ, TU, and EX as Cylon Basestars.

Message 6 of 13
Super Contributor

## Re: Utilization Intervals, Brackets, etc.

You asked about the older mortgage models, and whether they included EX 2.  Yes, EX 2 is the mortgage score for Experian.  It is an extremely old FICO model.  FICO released it in 1998 and therefore in FICO's nomenclature it is called FICO 98.

The other two mortgage scores (Equifax and TransUnion) use a slightly more recent model, released by FICO in 2004 (FICO 04).  As you see, however, even that model is still very old.

Each credit bureau makes its own tweaks to the model, and thus the EQ and TU mortgage scores behave differently, even though they were initially based on the same FICO model.  For example, the EQ mortgage score has a maximum possible score of 818, whereas the TU model has a maximum of 839, and in contrast to FICO 8 and 9 which has a maximum of 850.

The EQ mortgage model appears to be the most sensitive to number of accounts with a balance.  To give you an example, my EQ score was almost perfect, with a score of 817.  Then I had a recent inquiry and the number of accounts with balances went from 3 (two open loans and one card) to 10 (two loans and 8 cards).  My score dropped to 759!  It will drop even further when my new card appears.  Throughout that my CC utilization remained constant and very low (1-3%).

That shows you that TT is right when he says the mortgage models can react very strongly to number of accounts with a balance (plus a recent inquiry).

Bear in mind that these crazily old mortgage models will eventually be replaced.  Not in the next three months, but it's quite possible that three years from now the industry would be using something more recent.  Lenders' hands are basically tied by what Fannie Mae will permit, but there's been quite a bit of pressure to update to something more recent.  It's also quite possible that five years from now Fannie Mae will still only authorize the same ancient models being used today.

Message 7 of 13
Senior Contributor

## Re: Utilization Intervals, Brackets, etc.

@CreditGuyInDixie wrote:

Each credit bureau makes its own tweaks to the model, and thus the EQ and TU mortgage scores behave differently, even though they were initially based on the same FICO model.  For example, the EQ mortgage score has a maximum possible score of 818, whereas the TU model has a maximum of 839, and in contrast to FICO 8 and 9 which has a maximum of 850.

The EQ mortgage model appears to be the most sensitive to number of accounts with a balance.  To give you an example, my EQ score was almost perfect, with a score of 817.  Then I had a recent inquiry and the number of accounts with balances went from 3 (two open loans and one card) to 10 (two loans and 8 cards).  My score dropped to 759 It will drop even further when my new card appears.  Throughout that my CC utilization remained constant and very low (1-3%).

@CGID

This is a hidden gem of a data point: EQ Fico 04 817 => 759. It parallels my past EQ experiences: 765 => 809 (6/6 cards to 3/6 cards) and 809 => 764 (2/6 cards to 6/6 cards). In all cases aggregate UT was in the 2% to 5% range).

Do you have any corresponding data points on TU Fico 04 or EX Fico 98?

* Note 1: My TU Fico 04 score reacted more strongly to all cards reporting balances during the period 11/2017 - 11/2018 when I had an inquiry under 12 months on TU for a CLI. 823 => 812 (0 INQ); 823 => 801 (1 INQ)

* Note 2:  My EQ Fico 04 score reacted less strongly when lone INQ was removed with all cards reporting: 809 => 764 (1 INQ); 804 = 777 (0 INQ)

FWIW - below is the graph I generated from 2016 data updated to show impact on TU with an INQ < 12 months.

*** Data points on mortgage Ficos from other posters - going from 1 or 2 cards reporting to most/all reporting while maintaining low utilization would be of interest. ***

...

Fico 8: .......EQ 850 TU 850 EX 850
Fico 9: .......EQ 850 TU 850 EX 850
Fico 4 .....:. EQ 809 TU 823 EX 830 EX Fico 98: 842
VS 3.0:...... EQ 835 TU 835 EX 835
Fico 8 BC:. EQ 892 TU 900 EX 900
CBIS: ........EQ LN Auto 940 EQ LN Home 870 TU Auto 902 TU Home 950
Message 8 of 13
Established Contributor

## Re: Utilization Intervals, Brackets, etc.

@Thomas_Thumb wrote:

Aggregate utilization is weighted more heavily than individual card utilization or # accounts with balances on Fico 8. Number of accounts with balances has a relatively small effect on Fico 8. Unlike Fico 8, the older Fico "mortgage" models do put significant weight on # of accounts with balances - and presence of a recent inquiry may increase weighting further.

In your example above you are increasing #/% of revolving accounts with balances and this could cause a score drop of up to 5 points

Note what happens, report back and record the result for future reference.

@Thomas_Thumb , @CreditGuyInDixie , since you both replied with information about the effect of number of accounts with balances.

I only have an Experian update this time, moving from 1 to 2 accounts with balances, which is 100% of accounts in my case.

There was no drop in FICO 8 score! And I only lost 6 points on EX FICO 2, from 727 to 721. This is with both cards at 7% utilization, so aggregate at 7% also. 135/2000 (6.75%) and 453.41/6500 (6.98%).  Just 1 inquiry at EX from December 2018.

The only other changes are: ageing of 1 month to [ AoOA 1 yr 2 mo, AAoA 6 mo, AoYA 2 mo ], and a payment in full on the only card that had a balance.

-3 on EX Auto 8 to 680, -5 on EX Auto 2 to 698, -5 on EX Bankcard 8 to 698, +1 (?!) on EX FICO 3 to 713, -6 on EX Bankcard 2 to 726.

Both statement balances have already been paid, so now I can begin the 6/6/6% test to see what happens next month.

New CRADIS contacts, bearing 1 carom 10. Just think of EQ, TU, and EX as Cylon Basestars.

Message 9 of 13
Super Contributor

## Re: Utilization Intervals, Brackets, etc.

@CassieCard wrote:

I only have an Experian update this time, moving from 1 to 2 accounts with balances, which is 100% of accounts in my case.

There was no drop in FICO 8 score!

That's actually expected in most cases.  See my graphic below where going from 1 to all of my revolvers with a reported balance resulted in no impact at all to my EX FICO 8 scores.  The only time I ever saw any penalty was when all revolvers reported a 0 balance, thus leaving just my 1 (mortgage) installment account with a non-zero balance.

Message 10 of 13
Advertiser Disclosure: The offers that appear on this site are from third party advertisers from whom FICO receives compensation.