Capital one credit pull question

[Cblough93]

So capital one does their infamous triple pull, I believe they may be the only one who does this but there are plenty of other issuers who do a double pull. 

so my question is, let's say an issuer pulls 2-3 credit reports, but the information on the reports are widely different. How do they decide which report to use? Or do they use some kind of combination between the 2?

for example, my Equifax report is widely different from my other 2, there is certainly a large enough difference that it would cause an issuer to have very different opinions of my overall profile. So how do they decide which report to go off of?

[Anonymous]

When researching a subject do you (form an understanding/base an opinion/postulate a theory) on all the factual data acquired, or would you choose to treat huge amounts of it as nonexistent?

If you were a lender of money and you paid for 1-3+ credit reports (profiles):

• Would you only choose one of these reports/profiles to make your decision?
• Would you use all the reports/profiles to make your decision?
• Would you have no critical thinking of your own, ignore the data, and base your decision solely on the score itself?
• Would you have your own additional criteria to form your decision based on all factual data acquired?
• Would you stop at paying for only 1-3 reports or would you source even more reports/data/profiles such as Chexsystems, Innovis, LexisNexis, Teletrack, etc?

Maybe the more advanced question is.
What are all of the lenders secret internal algorithms and unmentioned criteria used to form their decision based on the data they acquire about you?
Good luck with that. I would love to know that as well.

----
Also: A good read
https://en.wikipedia.org/wiki/Critical_thinking 

[Cblough93]

@Anonymous 

I like where you are headed, Of course I always try to base my opinion off facts alone, but that isn't always easy. Not every decision is unbiased like that. Your personal feelings, opinions, and morals often get in the way of the facts. That's what makes us different from robots.

It's like people called in for jury duty but are released due to their personal beliefs. Like if the case we a drunk driving case, and one of the jurors lost someone due to drunk driving, They would not be able to judge that situation on facts alone, due to their personal feelings.  

If I was a lender I would probably pull 1 score and base the application off that. If I did pull 2, and the information was different I would use a combination of both profiles. 

but if I pulled 2 and the information was radically different. I think I would judge the application off the worse of the 2 profiles. Meaning I would really only be using 1 of the 2 profiles to judge the applicant. If 1 profile had a missing payment and the other did not, I certainly couldn't judge them off the better profile with no missing payments, that would be too risky. 

The trouble becomes deciding which profile is worse.

for example if I had an applicant, who had a TU profile with decent age metrics and credit history, very good util ratio, but had a 5 year old missed payment or two.

and he also had an Equifax profile with no missed payments, but it was a young profile, only about a year old, and it also had a vey high util ratio.

i would have to think about which profile i consider worse? The profile with missed payments, but great util? Or the young profile with little history, no missed payments but a lot of debt? My answer may be very different from yours. 

the facts are exactly the same, but my personal beliefs may be different from yours? Causing us to have different opinions on which profile is worse.

thanks for your response you gave me something to think about

[FicoMike0]

I have read that mortgage lenders pull three and use the middle score. I believe that Capone uses the mortgage scores for card apps. Maybe they use the middle score. There are statistical arguments for using the median vs. the mean. Of course, with a sample size of 3, it's all a crap shoot.

Many lenders look hard at recent accounts opened. I was turns down for that reason, with a score of 850.

[Anonymous]

Part of my point was they would use all of the information from all of the data they researched, compiled, collected.

And another part was that the "score" isn't or isn't completely relevant.

What matters is the data in the profile(s). And since the issuers are going to be using a critical thinking algorithm, whether it be human or human programmed computer, it will factor all data given in the decision. They are not going to pick and choose.

Except for denials. Then they will pick and choose from the worst line items. (Or just make something up)

[FicoMike0]

@Anonymous 

I'm not sure most banks do use all of the data and I do think the score matters. If they were really going to use all of the data, they wouldn't buy the score, they'd compute their own. Apparently nfcu does compute their own. 

The scores themselves are supposed to use all of the data to compute the probability of our being 90 days late in the next two years. When we see things like the ssl trick work, it looks like they didn't use all of the data. We're now hearing that fico10t and vs4 use trended utilization data. That data has always been available, they just didn't choose to use it in previous models.

[Anonymous]

---

@FicoMike0 wrote:

I'm not sure most banks do use all of the data and I do think the score matters. If they were really going to use all of the data, they wouldn't buy the score, they'd compute their own.

The scores themselves are supposed to use all of the data to compute the probability of our being 90 days late in the next two years. When we see things like the ssl trick work, it looks like they didn't use all of the data. We're now hearing that fico10t and vs4 use trended utilization data. That data has always been available, they just didn't choose to use it in previous models.

---

I do see what you are saying. You say you believe score matters. I say score isn't or isn't always completely relevant.

Lets use your above example.

"Many lenders look hard at recent accounts opened. I was turns down for that reason, with a score of 850."

So, your score of 850 was meaningless, and they didn't like some part or parts of your profile/report/data and denied you.

I have seen countless DPs like yours. Score was great. Profile solid. Denied for whatever reason they told you (or didn't actually tell you).

I have gotten approved for half of my cards at low 700 scores and one at about 685 that most of the savvy people here would have told me not to "waste" a HP and to wait for a year or "get your score higher". Oh, btw HPs don't matter.

My score(s) were meaningless, and they liked my profile/report/data.

I have seen countless DPs like mine.

One thing that all "scores" have in common. You are never told why you WERE approved.

[Anonymous]

Here is a new question.

Are credit card issuers legally required to pull/obtain a credit report?

No amount of search engine research (by me) returns an answer. And "Google Gemini AI" which is now occasionally at the top of google search results to give vague, misleading, or completely incorrect responses, and that has already repeatedly proven itself to me to be worthless, states "Yes, blah blah" and the cited link it gives literally contains absolutely no mention or reference to CC issuers being required to obtain a report.

The closest I can find is a Fannie Mae "Originating & Underwriting Selling Guide" which states mortgage lending requirements of having to obtain an In-File 3 bureau pull.
https://selling-guide.fanniemae.com/sel/b3-5.2-02/types-credit-reports 
This is apples to oranges.

[FicoMike0]

I think many banks have thresholds which trigger disapproval. For example its pretty clear that amex and nfcu trigger on your oweing them money, lol. It seems that Chase triggers on 5/24 new accounts. I think some, maybe penfed, trigger on 2/6 or 3/12 new accounts. If you hit a trigger, your score, or any other factor, doesn't matter. On the other hand, some of the same banks may trigger at a fico8 score of 700. Under that score, no credit, no how.

In these cases, they don't use all the data, they use only one parameter.

[Anonymous]

---

@FicoMike0 wrote:

I think many banks have thresholds which trigger disapproval. For example its pretty clear that amex and nfcu trigger on your oweing them money, lol. It seems that Chase triggers on 5/24 new accounts. I think some, maybe penfed, trigger on 2/6 or 3/12 new accounts. If you hit a trigger, your score, or any other factor, doesn't matter. On the other hand, some of the same banks may trigger at a fico8 score of 700. Under that score, no credit, no how.

In these cases, they don't use all the data, they use only one parameter.

---

@FicoMike0 

I completely agree with everything you said. But they don't need to use any of the rest of the data because one single data point was the reason for the decline. The same would be said if 2, 3, 4, 5 data points combined were the reason for the decline. But they all have one thing in common, the decline was based on the profile/data/report.

With one exception: your example of "may trigger at a fico8 score of 700 or xxx" (which I do not discount the possibility)

BUT, here is a DP from a quick search:

DW approved for AMEX Gold Card, Pre-qualified and then Approved w/ 630 EX FICO

https://ficoforums.myfico.com/t5/Credit-Card-Approvals/DW-approved-for-AMEX-Gold-Card-Pre-qualified-... 

That approval was profile/data/report driven and the "score" was meaningless.

Edit: I just reread that link above and the DPs posted show his DWs profile looked like *poop emoji*

------------------

I also need to clarify I am only speaking of score in the realm of unsecured revolving credit.

My understanding is that "Score" + profile/data/report is absolutely important in at least mortgage scores and in obtaining the best APR rates.

But for me, and anyone else, that finds the APR on a credit card meaningless ( @ptatohed ) APR of 9%,17%,25%,30%,36%,58.34% all = Zero