Does anyone know more about how FICO Score 9 differs from Score 8 model?

[Revelate]

Holy crap, I'm Gold Plated (tm) on FICO 9 for Equifax.  Previously Equifax hated me with a passion, now, hrm.

I'm nowhere close on EX/TU though admittedly they have recent dirtiness in them whereas Equifax does not (inquiry on EX, 30 day stupid stupid stupid late on TU).  I haven't really dug into this hard yet as balances are all over the map currently, but just as a comparitive snapshot.  FICO 9 appears to slap recent issues even harder than FICO 8 did potentially, but I still can't explain the difference albeit it looks tremendously much more like VS 3.0 as TT and I once debated a while back.

Just classic scores though the industry options look similar:

FICO 8: EQ 701, TU 691, EX 711

FICO 9: EQ 780, TU 707, EX 706

Common negative factors: Public record/collection (#1 on all), short credit history (#2, #2, #3 in same order as above), installment loan balances (#3, #4, #4).

Non-common negative factors:

EQ - Too many balances on credit accounts (#4)

TU - Recent missed payment (#3)

EX - Recently looking for credit (#2)

If you want more details I can dig into them in terms of differences, but that's a striking disparity when my EX is nearly identical to my EQ with the exception of that inquiry from two days ago.  My state tax lien is still on all of them, as is my paid collection, as is the old 30/60 day lates, still have a slew of recent accounts (<6 months) on all, and AAOA is pretty close to the same (1 month difference both on my calc and what the 3B report suggests).

[SouthJamaica]

---

@Revelate wrote:

Holy crap, I'm Gold Plated (tm) on FICO 9 for Equifax.  Previously Equifax hated me with a passion, now, hrm.

I'm nowhere close on EX/TU though admittedly they have recent dirtiness in them whereas Equifax does not (inquiry on EX, 30 day stupid stupid stupid late on TU).  I haven't really dug into this hard yet as balances are all over the map currently, but just as a comparitive snapshot.  FICO 9 appears to slap recent issues even harder than FICO 8 did potentially, but I still can't explain the difference albeit it looks tremendously much more like VS 3.0 as TT and I once debated a while back.

Just classic scores though the industry options look similar:

FICO 8: EQ 701, TU 691, EX 711

FICO 9: EQ 780, TU 707, EX 706

Common negative factors: Public record/collection (#1 on all), short credit history (#2, #2, #3 in same order as above), installment loan balances (#3, #4, #4).

Non-common negative factors:

EQ - Too many balances on credit accounts (#4)

TU - Recent missed payment (#3)

EX - Recently looking for credit (#2)

If you want more details I can dig into them in terms of differences, but that's a striking disparity when my EX is nearly identical to my EQ with the exception of that inquiry from two days ago.  My state tax lien is still on all of them, as is my paid collection, as is the old 30/60 day lates, still have a slew of recent accounts (<6 months) on all, and AAOA is pretty close to the same (1 month difference both on my calc and what the 3B report suggests).

---

Between what FICO says, what Thomas_Thumb posits, & my experience, these seem to be differences between FICO 9 & FICO 8:

-paid collections not counted (FICO)
-medical collections given less weight (FICO)
-rental history counted (FICO)
-recent inquiries/young age of accts given less weight (T_T) (consistent with my experience)
-smaller no. of accts reporting balances given more weight (T_T)
-premium for open as opposed to closed instalment loans reduced (T_T) (consistent with my experience)

[Thomas_Thumb]

Some poster data also suggests Fico 09 may not ding score as harshly for high balance to loan ratio on open installment loans.

No hard data on this though.

[Revelate]

---

@Thomas_Thumb wrote:

Some poster data also suggests Fico 09 may not ding score as harshly for high balance to loan ratio on open installment loans.

No hard data on this though.

---

Hard to say, my EQ FICO 9 is stupidly high whereas my FICO 8's and EX/TU FICO 9's are all in line with each other with a wretched loan ratio on open installment loans (7 months into a 15 year mortgage, gotta get comfy).

I don't intend to really stress test FICO 9 anytime soon as it's a paid pull for every data point and it doesn't matter to me nearly as much as the run up to the mortgage sequence of pulls I did.  

I can't fathom the difference between EX and EQ though, yesterday's inquiry is the only significant difference and FICO 9 was supposed to be closer together... I'm not going to take a random EQ inquiry just to see, but I guess the big question is what'll happen when that inquiry passes say 3 or six months: that wasn't a FICO breakpoint previously, but it's #2 on the reasons my EX FICO 9 sucks, so maybe it's a much heavier penalty than previously.  

[ch8ngeurstars]

My FICO 9 scores are 30 to 60 points lower than my FICO 8 score!!! I have multiple paid collections, more than 10. I'm not proud of it but I thought for sure my fico 9 scores would sky rocket and instead they're going to screw me once companies start using them. It was really discouraging to see those scores today. It almost seems like everything that I have learned on this site to make my scores higher wo't work fro FICO 9....smdh.

[CreditDunce]

---

@ch8ngeurstars wrote:

My FICO 9 scores are 30 to 60 points lower than my FICO 8 score!!! I have multiple paid collections, more than 10. I'm not proud of it but I thought for sure my fico 9 scores would sky rocket and instead they're going to screw me once companies start using them. It was really discouraging to see those scores today. It almost seems like everything that I have learned on this site to make my scores higher wo't work fro FICO 9....smdh.

---

I don't mean to pry... but does your CR also have multiple lates?  Especially the 90+ day late kind?  We know that paid CO's are not supposed to count, but we dont' know if FICO has upped the pain and/or time of pain for lates.  We are going to need a lot more data before we know how FICO 9 behaves. 

[Thomas_Thumb]

CD - Good question

I've read that Fico 09 is more lenient toward isolated 30 day lates with some age on them relative to Fico 08. On the flip side, Fico 09 is said to extract a progressively greater penalty with 60, 90 and 120 day lates - particularly if they are relatively recent. See link and associated exerpt pasted below.

https://www.autocreditexpress.com/blog/limited-credit-and-fico-score-09/

The Good News for Bad Credit Borrowers

The good news for consumers with credit issues is that the new FICO score will no longer penalize borrowers equally for 30, 60, 90 and 120 day late payments and will adjust the penalty accordingly. In other words, a 30-day late payment will have less of an impact on a credit score than a 60-, 90-, or 120-day late payment.

So buyers with shorter late payments should see their credit scores rise, while the process also introduces a way to better estimate the ability to pay for those borrowers with very little information in their credit files. This improvement, by the way, is especially important to consumers with poor credit and could affect their ability to qualify for a poor credit auto loan.

Add treatment of lates to the list of differences.

[CreditDunce]

Thanks TT.   I always assumed FICO 8 treated lates differently based on their age.  I learn something new everyday.

[Thomas_Thumb]

Consider the info in my prior post with a grain of salt - given the link source.

That being said, my takeaway is that Fico 09 amplifies the score differences among lates. Also, and (per other articles) Fico 09 may taper impact on score based on how long ago the late took place more so than does Fico 08.

Note: To the best of my knowledge Fico 08 does, per Fico communications, treat 30 day lates less severly in scoring than 90 day and 120 day lates.

[Revelate]

---

@Thomas_Thumb wrote:

Consider the info in my prior post with a grain of salt - given the link source.

That being said, my takeaway is that Fico 09 amplifies the score differences among lates. Also, and (per other articles) Fico 09 may taper impact on score based on how long ago the late took place more so than does Fico 08.

Note: To the best of my knowledge Fico 08 does, per Fico communications, treat 30 day lates less severly in scoring than 90 day and 120 day lates.

---

To my knowledge as well, also it held under Beacon 5.0 when that was the Scorewatch version where 30 day lates were suggested based on large amounts of anecdotal evidence to not factor after 2 years, though FICO 8 seemed to treat 90 day ones similarly to 30 day ones when we were talking 6ish years old based on some reports whereas FICO 04 was "known" not to.

A question I have, did FICO change their policies to try to limit application sprees by massively penalizing very recent inquiries?

I do have 6 scoreable inquiries on EX vs. 3 on EQ but the scores are 74 points apart (706 vs 780 respectively, which is night and day) and everything else is line item identical assuming that the algorithms are interpreting the tax lien status nomenclature the same.  In this case the last inquiry on EQ was 8/23/15, and EX was 1 day before the pull (3/2/16) with the rest being 8/23/15 or earlier in EX's case.  

I can't find any other difference after going through the EX base report, haven't done similarly on EQ but CK's data is equivalent for reporting statuses with EX.