VS4 vs VS3

[Trudy]

For those that get a VS4 (Synch or otherwise) have you found a difference between the 2?  With Synch now offering VS4 I'm curious regarding the DP's as this may be gaining ground.  For me, I thought the VS3 was different but not sure I can synch up the change "date" based on where and when the scores are provided.  In the last 3 months it seems to track the same for me.

From what I've found there are slight differences in weight although I haven't seen it reflected in my score comparisons, yet.  They have remained the same.

https://www.creditkarma.com/advice/i/new-vantagescore-4-0-explained/

I don't have the derogs that the link shows differences between the 2.  But for trended data, I've not been close to the debt I've had in the last 7 months as I've had nearly 3 years prior (much less) which leads me to question the look back period for trended data on VS4.  

Outside of the look back period, I guess I'm curious about how the different weight impacts the noted derogs between VS3 and VS4 for those that have them and have access to both scores?

[Anonymous]

I actually updated my VS4 yesterday morning after I got my BK off TU via EE. 

VS3 gained 23 points to hit 724

VS4 gained 10 points to hit 714. 

I have found in the few months Sync has been using it that VS4 runs higher than VS3 by a few points but within margin of error however it does surprise me that VS4 basically shrugged at the loss of such a large negative. 

[Anonymous]

I can’t speak of the derogatories, but the trended data is going to go back 24 months because that’s what the CRAs keep on payment history from my understanding.

[Anonymous]

If the trending data involved some sort of averaging, which the whole concept of trending would imply is almost necessary, that would explain the slighter response to the dropoff of @Anonymous's BK - it would show 23 months with a BK and one month without. It would stand to reason that as his BK fades into the past, the gains seen in other scoring models would be realized over time with VS4.0 rather than more immediately like in the other models.

VS3.0 has always been on drugs for me with respect to how it responds to changes (ie, get rid of a derog, age accounts past known thresholds, and decrease balances, and as a result my score falls), but at least I can predict its screwiness somewhat. With VS4.0, which I only see in my Lowe's account, I see fairly big jumps in the 20-40 point range sometimes overnight, and even a drop of 20 one day followed by a bounce-back the next day, when I have no alertable changes in FICO, no aging thresholds, no inquiry drop offs, and no accounts reporting balance changes. I'm truly at a loss as to how it gets these numbers when in the VS page in my Lowe's account, it shows the biggest factor being utilization and that hasn't appreciably changed despite the score fluctuations.

[GApeachy]

4 months ago when I got my first VS4 TU from Care Credit my score was (Oct., 2019) 758 while CK V3 Tu shows 651. 

For January 2020 the gap closes. VS4 695 and CK VS3 693.

I just looked at both of my scores today and they are both 690.

eta: forgot to add, I have one 30 day late from 2016.

Also what's weird to me is my tcl went up, debt went down a lot. (have always pif, except for loc but pay more than double required)

[Anonymous]

I sure hope you're wrong @Anonymous because if they did that, they have effectively made bankruptcy a 12 year affair when it was already a 10 year one. Let's hope FICO didn't make such a dumb move in 10 T. 

[Anonymous]

I hear ya. I wouldn't think it's done in a linear fashion and the full recovery would take two years from the point it drops off. It may be that the score effect fades starting sometime before the falloff. No way to know yet since it's fairly new and its adoption is very limited with only Synch using it. Right now all we have is supposition and guesswork. The difference in points gained is more likely a function of the overall changes in the way 4.0 looks at an entire profile, not just one trend factor.

[Anonymous]

VS4 can only take into account what is on the credit report. Trended data, from my understanding, is an evaluation of payment history.

Payment history goes back 24 months. If a bankruptcy dropped off, it’s no longer on the report and therefore cannot be considered. It doesn’t have a payment history to go back 24 months on, so it can’t be stretched another 24 months. Where would the algorithm get the information from if it’s no longer on the credit report?

Let me give an example if you close an account, payment history is still there for 24 months. In contrast, if an entire TL is deleted, that payment history disappears.

Bankruptcy, when it runs it’s course, is removed from the report just like deleting a TL. If the data is not there, the algorithm can’t consider it.

Just my two cents.

[Trudy]

Very interesting feedback.  The information for VS4 states "historical UTL rate and payment info (trended data)".  Birdman7, are you saying you believe the trended data is only for payments since the reports shows past payment history and not past UTL?  From the trended concept I assumed the scoring model somehow has access to this information even if it doesn't appear on the report.  Maybe we'll start seeing this on our credit reports?.?.

[rbentley]

My VS 3.0 and 4.0 scores are virtually identical.