Credit in the News

[pipeguy]

FICO's new scoring model may make it easier for some to buy a house or get a credit card. But major lenders have been slow to switch, and the entire mortgage industry has yet to sign on.

Complete story: http://money.msn.com/credit-rating/new-credit-score-could-help-millions-credit.aspx?page=0

Kind of interesting story on FICO scoring, both new and old.

[Revelate]

It's make it harder for some others to do similarly.  Until the GSE's and FHA switch to the '08 or some other future version of the FICO models, nobody else will either.  

It wouldn't surprise me in the slightest if we actually see a new model basis sometime in the next few years either from the data when the recession really started digging in; however, it takes so long to get anything done in Fannie / Freddie / FHA, and then getting all of them to agree, we may still be using '04 scores in 2044.

Does anyone recall or otherwise know why the mortgage industry stepped up to the '04 scores to begin with, and when that happened?

[pizzadude]

I love this quote ~  "For individual consumers, it depends," Quinn says. "For some consumers, their FICO 8 score will be higher than their previous score, some consumers will be lower, and others won't change at all."    REALLY ?!??!

For banks there isn't any incentive to switch to a newer scoring model, and it would likely cost them large amounts of time and money.   If there is a business case that the newer model could better identify more profitable customers maybe it would make sense.   

[Revelate]

---

pizzadude wrote:

I love this quote ~  "For individual consumers, it depends," Quinn says. "For some consumers, their FICO 8 score will be higher than their previous score, some consumers will be lower, and others won't change at all."    REALLY ?!??!

For banks there isn't any incentive to switch to a newer scoring model, and it would likely cost them large amounts of time and money.   If there is a business case that the newer model could better identify more profitable customers maybe it would make sense.   

---

I'm certain they're simulating it.  If it shows that FICO 8 provides better resolution than FICO 4 (and therefore less mortgage defaults), we'll switch virtually overnight, but until then you're correct, there's no compelling need to do so.

It probably wouldn't be that much time or money, technically it's an easy change (no new infrastructure / capital to be spent), it's just a different algorithm.  The problem is coordinating that industry wide, though they'd almost certainly do it with a phased approach / grace period for mortgage lenders.  90 or 180 days where they'd accept either score set for example.

[daboss28]

I find the part interesting about Citi cards adopting the model in many products
as of June 2011. When the score they gave me as of 2/2012 ... a good eight months
after this proposed layout was clearly not a FICO.

[Revelate]

---

daboss28 wrote:
I find the part interesting about Citi cards adopting the model in many products
as of June 2011. When the score they gave me as of 2/2012 ... a good eight months
after this proposed layout was clearly not a FICO.

---

They almost certainly did both: a FICO and also an internal.  Which one they give you I don't know is actually mandated anywhere.  I do know a company I have a loan product with (since I used to work there) provided me a FICO, but also based most of their underwriting decision off an internal algorithm they came up with.

Now that I think about it, it wouldn't surprise me at all if multiple algorithms were run even of the FICO variety: the lenders have a tremendous amount of data available to them in terms of historical payment history, and they can look at the score whatever model gave them and see how that trends over all of their customers.  I'm certain they do that by default with their internal scores / whatever pull they do make, but there's no real reason they couldn't run FICO 4 and 8 side-by-side and see which worked better over time.  The incremental cost to the additional score is probably minimal, vs. getting even a 1% better lending resolution rate, which is huge when we're talking about the aggregate money each has in the field, and breakeven is probably a lot less than 1%. maybe by more than an order of magnitude.

[teenastie]

The most interesting part of this article did not come from the writer but from the comments! There are many many misled, uneducated, bitter, conspiracy theorist out there... I was one of them... Before I found this site

[pakman92]

I don't see how this could 'help millions.'  regardless of the algorithm, essentially the produced score is on a curve. only a set percentage of people will get 760+.  it may move some people from another, but it wont change the distibution.  so while it may 'help' millions, it will 'hurt' the same number of folks...