"Credit Mix" maxed with just 2 account types?

[Anonymous]

Anyone else find this to be kind of goofy?  It just seems strange to me that this category of scoring really only has 2 possible outcomes... credit mix being unsatisfied, meaning you have only 1 type of account on your credit report, OR satisfied, meaning you have 2 (or more) types. 

Then there is no added benefit of having any more than 2 types.  I get it that there are a billion different types of installment loans.  A mortgage to me is quite different than most, simply based on the sheer length of time relative to many of the other types.  There are also different types of revolvers. Then charge cards.  It would just seem to me that if 2 types is "better" than 1, wouldn't 3 be "better" than 2?  Under FICO scoring, the answer is no.  I just find this category to be kind of bogus in terms of the way that it's scored.  Granted, it's not a huge category or anything, I just feel like FICO could have done a bit better here. 

I don't know, maybe I'm alone on this.  But, to me, someone that's got 1 credit card and 1 auto loan being scored the same [with respect to only Credit Mix] as someone with 1 credit card, 1 auto loan, 1 mortgage, 1 charge card, 1 HELOC and 1 student loan just seems a bit silly.

Anyone else have an opinion on this?

[Anonymous]

This is a hard thing to test definitively.  I cannot be sure that there is no scoring benefit (traceable solely to credit mix) from any account subtypes.  I am all for someone with a very stable profile and exactly one mortgage loan (no other installments) implementing the SSL technique -- or a person with zero charge cards adding a true classic charge card -- just to see what  happens.  If the test is extremely well controlled and there is a boost, that would seem to indicate a benefit, but if there is a score drop or nothing, that might be because of the change in AAoA or Age of Youngest or an Inquiry.

As a practical matter I try to advise against people doing it because such people are typically frantic people trying to get extra points for some specific need (e.g. impending home or car purchase) and they are not in a position to philosophically take a scoring hit for the sake of science.

Also the only model family we can test with even remote accuracy is FICO 8 -- not the mortgage models or FICO 9.

Testing is further complicated by the possibility that some of the extra points might be gained via having accounts of different types that were or are open for a while -- TT has conjectured this I believe.  (A "while" might mean a year, say.)

As a practical matter, the credit mix question seems to reduce to this advice:

      * open the accounts that are right for you, ignoring any phantom benefit from mix....

      * unless you don't have a few credit cards and an installment account.  (If no installment account, implement the SSL technique.)

The theory part is harder to chase down.

[Anonymous]

Credit Mix seems pretty good to me, actually.  Because lenders only have a snapshot in time (other than the trending data products), they want to see a person's current ability to use and manage credit well, and make sure there's no past history of deliquency.

Credit Mix makes sense to me.  If your credit card usage is $0, your "utilization" portion of your score is great, but your "credit mix" is poor because you're not using credit cards when that day's report is pulled.  Same with installment loans -- if you have $0 balances there, a lender can't see if you are managing loans well TODAY.  The mortgage score's negative points versus a non-mortgage loan also makes sense to me -- a mortgage is the most "secured" loan one can get, and people typically will pay their mortgage even if they stop paying everything else, so it shouldn't count as much on a credit score as other installment loans.

Giving a ding to Credit Mix for "consumer finance company" also makes sense, since these CFCs are incredibly predatory many times and the people who sign up for them typically either are subprime borrowers, desperate borrowers or borrowers who haven't proven themselves with prime credit products.

I get Credit Mix -- since the lender only sees a snapshot, they have to calculate a real risk factor based on what they see this one time.

[Anonymous]

I'd like to hear more opinions on whether or not anyone believes that a 3rd credit type would offer any benefit with respect to scoring in either the short or long term.  I personally don't believe it does.  I figure if it did, someone would have reported something about it at this point. 

I do get what you're saying that it would take a very stable and just the "right" profile to even test it.  I do however believe that the benefit would happen quickly, not over time.  My reasoning for that is that if you implement the SSL technique, the benefits are realized right away once it reports.  I would think that if there were any added benefit from other installment loan types it would be seen right away.  This can more or less be verified by adding any type of installment loan to a profile that has no installment loan, not just a SSL.  I know people have tested adding a charge card to a file of just revolvers and no one has reported (that I know of) any scoring benefit from doing so. 

I could see there being benefits with a mortgage related to time, as it's sort of a unique type of installment loan that's quite different from the rest, but of course that would be next to impossible to test.

By the middle of next year the two auto loans on my file will be gone (one in Feb, the other in June) which will leave me with just a mortgage.  I'd have no problem adding a SSL at that time for the sake of testing.

[Anonymous]

I think when those autoloans drop off, you will see a small ding -- mostly because some of the profiles I've analyzed did see the ding (sometimes 10 points, sometimes 15 points, a lot depends on the age of the mortgage).  The Alliant SSL will return those points with immediacy.

My personal reference that I use for forecasting is based on analyzing many credit profiles, but not enough to feel secure in my forecasts.  I've come up with the following "rough" numbers:

Credit Mix (Credit Cards, minimum 1 card must show activity or no points offered at all)

1. One credit card = 5 points
2. Two credit cards = 8 points
3. Three credit cards = 10 points
4. Four credit cards = 13 points, assuming aged 24 months for full value
5. Five+ credit cards = 15 points, assuming aged 24 months for full value

Credit Mix (Installment Loans, minimum 1 loan must show activity or no points offered at all)

1. Has a loan product, none are a consumer finance account = +10 points (additive)
2. Does have any type of loan, mortgage or non-mortgage = +15 points (additive), assuming aged 24 months for full value
3. Has a non-mortgage loan = +15 points (additive)

Between these two, the credit mix = 55 points, which is 10% of FICO possible points.  Naturally there's no way to test this, but I am constantly tweaking things based on profiles people send me and I feel more and more confident with these numbers.

Again, this is just wild conjecture on my part as I work towards writing a working FICO simulator.  It fits my own profile very well based on watching new accounts show up this year during my rebuild.

[Thomas_Thumb]

---

@Anonymous wrote:

Anyone else find this to be kind of goofy?  It just seems strange to me that this category of scoring really only has 2 possible outcomes... credit mix being unsatisfied, meaning you have only 1 type of account on your credit report, OR satisfied, meaning you have 2 (or more) types. 

Then there is no added benefit of having any more than 2 types.  I get it that there are a billion different types of installment loans.  A mortgage to me is quite different than most, simply based on the sheer length of time relative to many of the other types.  There are also different types of revolvers. Then charge cards.  It would just seem to me that if 2 types is "better" than 1, wouldn't 3 be "better" than 2?  Under FICO scoring, the answer is no.  I just find this category to be kind of bogus in terms of the way that it's scored.  Granted, it's not a huge category or anything, I just feel like FICO could have done a bit better here. 

I don't know, maybe I'm alone on this.  But, to me, someone that's got 1 credit card and 1 auto loan being scored the same [with respect to only Credit Mix] as someone with 1 credit card, 1 auto loan, 1 mortgage, 1 charge card, 1 HELOC and 1 student loan just seems a bit silly.

Anyone else have an opinion on this?

---

Although one can clearly achieve a Classic Fico 8 score of 850 with only revolving credit and one type of installment loan, that does not mean points for that category are fully allotted. I receive reason statements from time to time suggesting I am leaving some bonus points on the table for not having an auto loan either open or closed on file. Perhaps that is the primary reason I can't get to 900 on Fico 8 enhanced for Auto but can on Bankcard.

Side notes:

1) I continue to believe a true charge card is a separate mix category from revolving credit cards. Certainly charge cards are treated differently in the various Fico algorithms relative to aggregate utilization.

2) In addition to account categories, QTY of accounts is a factor. I tend to agree that 1 installment loan + 3 revolving cards is sufficient for achieving a Classic Fico 8 score of 850 given sufficient age and proper account management. However, some bonus (buffer) points are likely being left on the table. 

[Anonymous]

TT, what sort of research have you done or what leads you to believe that charge cards are considered differently than revolvers with respect to FICO scoring?  I'm interested to hear about that.  Also, with charge cards, I would assume that if their presence helped it would be exactly that, their presence, as from what I understand you can't really report a balance right?  Or can you?

[Anonymous]

---

@Anonymous wrote:

TT, what sort of research have you done or what leads you to believe that charge cards are considered differently than revolvers with respect to FICO scoring?  I'm interested to hear about that.  Also, with charge cards, I would assume that if their presence helped it would be exactly that, their presence, as from what I understand you can't really report a balance right?  Or can you?

---

Hi BBS!  Charge cards certainly do report a monthly balance, like pretty much every other revolving or installment account.  Indeed, way back in the day, they even factored into utilization -- the card's highest historic balance was used as the card's credit limit (for util purposes).

But regardless reported balances don't apply toward the Credit Mix category (the subject of your initial question).  The amounts that accounts report affect FICO's "Amounts Owed" category. 

[Anonymous]

So, the way we have "revolving" utilization and "installment" utilization, do we have charge card utilization?  Is that utilization your current reported balance against your highest reported balance?  If that's the case, would the best move be to report a really high balance once, then report small balances from there on out so that utilization on that charge card is always low? 

I've considered adding a charge card to my portfolio for probably a year and a half now, about the amount of time I've been a part of this forum.  The only reason I considered it though was for the purposes of potentially improving my score.  I don't see any real practial use for one [over a CC] but then again I know nothing about any potential perks that may come with a charge card over a CC.  If anyone has additional insight on this topic, I'm all ears!

[Anonymous]

---

@Anonymous wrote:

So, the way we have "revolving" utilization and "installment" utilization, do we have charge card utilization?  Is that utilization your current reported balance against your highest reported balance?  If that's the case, would the best move be to report a really high balance once, then report small balances from there on out so that utilization on that charge card is always low? 

---

If a credit scoring model includes the charge card in its revolving utilization, as FICO 98 and FICO 04 did (and may still do) then there is not a separate kind of utilization for it.  Rather those old models just used the highest historic balance you ever had as a proxy for your credit limit.  Otherwise they treated it like a credit card.

FICO 8 and 9 by contrast ignore charge cards when calculating utilization.  The only caveat to the "ignoring" claim I just made might be if it is a charge card that permits you to treat it like a credit card (and carry a balance).  With certain Amex cards that's an option.  It's conceivable to me that FICO 8 detects that and then somehow includes it in the utilization.  Thomas Thumb would likely know the answer.

Bear in mind that almost every charge card has an annual fee.  Unless a person is happy paying that annual fee, then a charge card is a winning proposition only if you get it for 11 months, get a huge sign up bonus, and then cancel it.  Paying a fee for a possibly phantom scoring bonus strikes me as not worth it.