Why does the same new account provide different scores at different Credit Bureaus

[Anonymous]

Using the FICO monitoring service, I just saw that my new mortgage refinance account I opened in February hit 2 of the 3 credit bureaus this week (looks like the 3rd is still pending). But in one bureau the score went up by 13 to 711, at the other it went up by 9 to 694. I realize the system isn't alawys 1 for 1, but one would imagine that for something as standard as a 15 year mortgage, FICO and the CB's should be able to hit the same number within 1 point or thereabouts. I mean the difference between 9 and 13 is almost 50%! Anyone know why this makes sense?

[Revelate]

---

@Anonymous wrote:

Using the FICO monitoring service, I just saw that my new mortgage refinance account I opened in February hit 2 of the 3 credit bureaus this week (looks like the 3rd is still pending). But in one bureau the score went up by 13 to 711, at the other it went up by 9 to 694. I realize the system isn't alawys 1 for 1, but one would imagine that for something as standard as a 15 year mortgage, FICO and the CB's should be able to hit the same number within 1 point or thereabouts. I mean the difference between 9 and 13 is almost 50%! Anyone know why this makes sense?

---

Welcome to the forums!

The algorithms are customized for or by the bureaus.  FICO scores are not 1:1 between the bureaus, which has been both a consumer and a lender complaint over time.

I don't know why it's not one single algorithm for a given release, but it simply isn't unfortunately.

[Anonymous]

Plus your 2 scores are different, so there is most likely something different between the two reports, so the mortgage may well have affected them differently even if they hadthe same actual algorithm.

[Anonymous]

---

@Anonymous wrote:

Plus your 2 scores are different, so there is most likely something different between the two reports, so the mortgage may well have affected them differently even if they hadthe same actual algorithm.

---

Thank you both.

I considered this option and initially thought it unlikely because if the exact same (good) data point were to have different effects, it should have a higher point effect on the lower score (using something like relative weighting) otherwise the algorithm would never converge. Unless of course the "worse" data we are talking about is some sort of temporal check on how high you can go, like a bad entry at some recent time that tells every good entry in the report to deduct 40% and is holding everything down until that a prescribed period passes. The entry is in the lower score, but not in the higher score. Now resolving, or at least knowing very specifically what is causing this kind of thing is what I would pay cash for, I don't myself see big differences in the two reports that should make this so. But what FICO and other monitors offer is bland high level stuff. I guess that is what has to be if they make money on both sides of the coin. 

OTOH, the idea that they have different algo's for different credit bureaus is commerically appealing. That is how I would set up the system too if I were told industry participants needed to make good money in the business. Making it easy for customers whether merchants or consumers to comparison shop by standardizing things isn't very good for providers in any industry.

[Anonymous]

If you have the one account that is completely clean (no negatives) it is likely in a different bucket, while the other one may be in a bucket with derogatory marks, so you may be capped point wise in certain aspects. Before my last baddy came off experian (didnt appear on TU or EQ) every time i had positive things i got less points for them evrn tho ex was already the lowest of the 3. However when the last item came off, EX jumped above the other two. My file is thin so changes are probably slightly magnified but its obvious that the underlying data is the same as there is very little to sift through. 

If you look at the reason codes (you should get 4 of them) they may give you some clues as to why the scores are different. I also think that your expectation of convergence at a certsin point may be off. Very seldom have even 2 of my scores been the same let alone all three, despite my thin file and the data being the same. I have a variance of almost 25 points between TU and EX currently. Eq sits almost dead center between the two. Toss in the fact that EX updates waaaaay faster than TU and 3 of my 5 cards report the same day each month and for a 2 day period sometimes I have closer to a 40 pt gap. Then TU updates and they come closer again. Obviously at extraordinarily high scores there will be a convergence, but for those of us hanging out in the middle, something simply being coded wrong or data missing for a month or something like that may push or pull scores in different directions. 

I had a month where 2 cards showed a balance and TU hit me for 15 points for it. EX didnt move. Eq fell a couple. Toss in various amounts of inquiries that cost a variable amount of points and any number of permutations is possible. My last mortgage Inq cost me 4 on EX 2 on TU and 0 on EQ, despite total inq being relatively close (10/10/8 last 2 years 6/6/6 last 1 year 2/2/2 last 6 months).