Large difference between agencies

[Tuscani]

---

denbar2003 wrote:

---

Tuscani wrote:

Remember, there are different versions of FICO for each CRA.

---

But ... but ... WHY is it necessary that they have different versions for each agency?  Why don't all the agencies have the same model?

---

Remember, each CRA is it's own entity. They have nothing to do with one another in terms of the data they maintain. Thus, the reason Fair Isaac had to develop difference versions was probably do to different needs and requirements. Nonetheless, I think there should be one CRA and one formula!

Message Edited by Tuscani on 11-28-2007 03:45 PM

[haulingthescoreup]

Here is my theory: each CRA has different credit behaviors that it values and dislikes. So each one has requested a formula that will reflect those preferences. Everything I have seen shows that EQ penalizes lates, and look at my scores. Lates are my only baddies. The reports are virtually identical. TU likes mortgages, and fewer than half of TL's reporting balances. EX seems to dislike inq's (little shaky on that one, LOL.)

My three young adult kids each have absolutely identical credit reports (for each kid, I mean --EQ=TU=EX.) Yet each have different credit scores, and the differences reflect what I said above.

It makes sense--the CRA's need to market themselves to creditors (their customers.) So if they say that they have a unique spin --they look hard at lates, or util, or inqs-- that will give them a distinct "personality" for potential customers.

[Dawn]

---

haulingthescoreup wrote:
My three young adult kids each have absolutely identical credit reports (for each kid, I mean --EQ=TU=EX.) Yet each have different credit scores, and the differences reflect what I said above.

---

I'm glad that someone has such a "clean" set of reports to draw conclusions from.  Comparing my three CRA reports to one another ... there is no apples to apples comparison.  There are several accounts that are reported on one or two of them that are not on the other(s). 

[haulingthescoreup]

That's why I badgered them into letting me pull their scores --for research purposes!

Of course, it was useful to get baseline scores for each of them, so that they can start focusing on their own credit. With such short credit histories, they can see exactly where they are, and what to be ready for over time. I was relieved to see that one has scores lower than mine! (That's terrible to say, I know --but she's only 19, so she will bounce back in 6 months.)

[llecs]

I contacted customer support on-line and asked them this question:

"When I pull my reports (all 3) and scores via MyFico, does MyFico use the same formula to calculate the three scores on MyFico? "

Their response:

"Yes, you get a FICO score when you place an order with myFICO."

Yesterday, I went to my neighbor and asked to borrow his hammer. He said, "Sorry, I'm out of chicken soup." He continued, "One excuse is just as good as another."

I know Fair, Isaac & Co. created formulas for each CRA and others too over the years. Wouldn't it make sense to get a patent on that formula and sell it? There's no economic sense why they'd create a new formula everytime. Unless...they never got the patent, back in the day, and EQ or whoever first got it patented it first. Then the others couldn't copy it and Fair, Isaac is tasked to do new ones each time, thereby resulting in 3 different FICO formulas. Just thinking.

[haulingthescoreup]

Well (and I'm still spinning my theory here), the CRA's are the actual entities that the lenders deal with. FICO is the mathematics wonk that has whatever data they have on borrower behavior and the ability to write formulas to describe risk of borrower default. So if CRA #1 wants to have a formula and score that predicts borrower behavior with special focus on 60+ day lates, more than 3 inqs in a year, and the presence of a mortgage, presumably they could present that to FICO, and FICO would create an algorithm that would penalize for lates 60-days or greater and having 3 or more inqs in a year, but reward for having a mortgage. Or whatever combination of factors that the CRA wanted.

I don't think that FICO is the great guru with all knowledge of human borrowing behavior. I think that they're more the propeller-beanie guys. (Sorry, Tusc!) I think it goes back to what the CRA's want. Maybe it is more FICO-driven, but this angle makes sense to me.

I keep going back to looking at my kids' reports, and the range of scores that each have across all three bureaus, and one CRA score whines about no installment loans present, and another gripe about too many accounts open with balances ("too many" in this case meaning 2), and EQ zapped DD#2's score for a one-year-old 30-day, but didn't blink at DS's 30-60-90-120 string from 2 years ago, and on and on.