Vantage scroes terrible

[Anonymous]

The newest Vantage model is terrible or else Credit Karma is broken. Credit Karma Vantage scores are 75 points lower on TU and 90 points lower on EQ compared to FICO. What gives? I thought Vantage was supposed to be "fairer" since it does not factor medical collections. I have one medical collection on EQ that isn't on TU or EX.

[tonyjones]

VantageScore is useless, my TU FICO 08 Score is 80 points higher than both the CreditKarma VantageScore for TU/EQ.

[sterlingrose1]

I am so relieved to see this post. I just got a free score at Credit Karma because I had two creditors agree to take derogatory information off my report, and I wanted to see if it was gone yet. Almost had a coronary when I saw that Vantage says my credit is "very poor," nearly 100 points lower than my FICO scores recently updated on this site. I'm trying to buy a house and 3 mortgage lenders have told me my score is what I see here on myFICO. So whatever is wrong with CK's Vantage scoring, at least it doesn't seem to bear relationship to the real world.

[DantGwyrdd]

Credit Karma is a great resource to stay on top on your credit report (where else can you get an updated report for free every week?), you just need to learn to not give too much credit to the V3 score it gives you.

[Anonymous]

Weird thing for me is that my Vantage scores are pretty close to my Fico scores.

TU

Vantage - 789

Fico - 796

EQ

Vantage - 733

Fico - 721

[travisr0]

Definitely best to try and se CK as a metric only to try and gauge similarities and patterns within your scores.

[Aarvard]

---

@Anonymous wrote:

Weird thing for me is that my Vantage scores are pretty close to my Fico scores.

TU

Vantage - 789

Fico - 796

EQ

Vantage - 733

Fico - 721

---

Mine too. TU 08 765, TU Vantage 768.

[Anonymous]

My vantage scores are similar to my fico 8 scores, actually a little higher.  The biggest difference between the 2 on my file is the AAoA.  Vantage doesn't seem to recognize the older closed accounts.

TU:  Fico - 682;  Vantage - 688

EQ: Fico - 685; Vantage - 700 (March 12th had it at 689)

My DH has higher Vantage scores than Fico but his are also within 5 - 12 points of each other.  His file has the same difference as mine.  The AAoA is way off and it appears not to recognize the older closed accounts.

[cashnocredit]

My VS3.0 scores have been pretty close to my FICO 8 scores. Sometimes a bit higher, sometimes a bit lower, but typically within 20 points. My VS3 scores are exactly 800 while FICO scores are about 790 right now, down from the 800s since I decided to let all my accounts report balances and current util is up from 1% last month to 7%.

Keep in mind that VS3 is radically different from FICO scores of any generation prior to FICO 9 which just came out and is not yet used.  VS3 scores ignore paid collections (but not paid lates or chargeoffs from OCs.)  VS3 also, and I believe this is quite important, uses payments made in addition to balances. This data wasn't available with previous generations (prior to VS3.0 or FICO 9) but FICO's banking analytics blog noted last year that payments near the minimum (which is also reported) indicate a significantly higher risk. I believe this factor is one of the reasons these new generations of scores can claim to be significantly more predictive than earlier models even though they completely ignore paid collections - a major improvement IMO. It was non-sensical that paid collections should be as negative as unpaid ones but VS3 and FICO 9 completely eliminate them yet still claim increased risk prediction accuracy.

[flan]

---

@Anonymous wrote:

My vantage scores are similar to my fico 8 scores, actually a little higher.  The biggest difference between the 2 on my file is the AAoA.  Vantage doesn't seem to recognize the older closed accounts.

---

Closed accounts, particularly old closed accounts, have very little predicitve power. I don't know if vantage 3.0 completely disregards them, or if it's just CK doing it, but I don't doubt that vantage weights them as very unimportant. 

Since Vantage claims VS3 has better predictive power than fico's scores, and has more granular data, and more data types, it would be very surprising to find that there's a general trend, like VS3 is usually lower than fico or usually higher.  At the low end and the high end, the scores will tend to be closer, but in the middle range (which is the interesting range for most lenders) there will a lot of variation.  Two files with the same fico score may show one that does something that VS3 considers risky, but which isn't risky to fico; while the other account shows positive attributes that fico doesn't track.  The risky account will have a lower VS3 than fico, while the not-risky account will have a higher VS3.  (This is true regardless of whether VS3 is actually a better predictor of risk, or not.)  Also, Vantage claims the VS3 score is harder to "game" than fico.  That may mean that they're using the payment data stream for that, or that they're watching past utilization, or somthing else.  That will also cause divergence with fico scoring. 

If VS3 is more accurate than fico's models, lenders will use it more heavily.  It's too early to tell, but I have no doubt that major lenders are doing the tests to see.