My wife's credit score via VantageScore 3.0 (via CreditKarma, which is based on TransUnion data) is 790.

In contrast, My wife's credit score via FICO® Score 8 is 670 (also based on TransUnion data).

My best understanding for what is driving this is that one of her credit cards tends to report data to TU a few days late (e.g. 35 days after monthend versus 30 days). VantageScore doesn't care about this, as it shouldn't, but FICO Score 8 does.

I called TransUnion, but they don't manage the score - they just track the data. Obviously I'll try to harass her credit card's bank to report faster, but I'd also like to know - is there anyone I can contact at the FICO score calculation group to sort this out? Who the heck would I even be calling?

  • 7
    Out of curiosity, why do the five days matter? – RonJohn May 21 at 22:44
  • 3
    Obviously I'll try to harass her credit card's bank to report faster Why not just cancel that card? If you think that's the problem, just stop the problem... – quid May 21 at 22:59
  • @quid There's no reason to cancel the card (and take a hit on your credit score) when all you need to do is have them change your billing cycle to better align with their reporting cycle. – user71659 May 22 at 4:55
  • 3
    @user71659, there is also no reason to keep the card. People spend far too much mental capacity letting their credit score wag the dog; I would just cancel it, you might change the cycle; potato potatoe, this is a big nothingburger, – quid May 22 at 8:06

For full context here is a related question you posed - How can I understand historical changes in my credit score?

One of the points I made there which, for the sake of posterity, I'll repeat here is that you can't reconcile the scores calculated using 2 different models. Because:

  1. Algorithm is different
  2. Service providers are different and operate at their cadence

Short answer: Don't stress about it

Not just the model but when they poll the credit card data and what the data shows at that time is arbitrary (but at a cadence).

For ex, credit cards by themselves will not post their data to the scoring application. The only way a scoring application (service provided for free by credit cards, credit monitoring apps like Credit Karma) can do the scoring is to decide when to poll all the accounts you have. On that date, whatever credit info (like utilization, age of accounts, new accounts opened, timely payments, loan, mortgage) is taken and a score is calculated. This may fall on a date that is different from the statement date or payment date of your cards.

Which means, if they had waited until you eventually paid the bill (even on time), they'd have seen close to 0% utilization. But then some other card having different cycles may have higher numbers. So all you can do is to use these tools to generally monitor the health, get an indicator of the scores to have an idea, correlated score changes with activities and report incorrect activities (not scoring cycles or reconciliation).

And when you actually apply for a new loan (auto/home) the score by a particular model maybe irrelevant because often, they take more factors into account like income, expenses, collateral, payment history (without assigning numbers). So, don't sweat it. Look for discrepancy in the activities and worry about fixing those with the bureaus/lenders.

Personally, I have 2 scores differing by 100 and some differing by 20s at any given time. Looking further it looks like some card info (as seen by one of the apps) is delayed by even 2 months. This is, likely, because the last time they polled there may have been an error, or the credit card returned the same info, etc. (Software wise). But the info returned by all of them are the same wrt the factors (utilization, balances, timeliness, age, etc)

  • 3
    Credit scoring applications don't poll accounts -- there's sampling in both directions, reporting accounts to the bureau, and requesting reports from the bureau. +1 for the main point that different models should not be expected to agree since they are estimating different things – Ben Voigt May 22 at 4:08
  • @BenVoigt It looked to me that, for ex, Credit Karma, polls credit accounts separately because some cards are delayed vs others. But that may just be unaligned reporting. Other finance tools like PersonalCapital (that may not calculate credit scores) do poll individual accounts. But my main msg was that these may not be aligned for various reasons and trying to align them is not necessary. The exact mechanism (not just scoring model) may differ and so it is not worth sweating it. – perennial_noob May 22 at 17:53

You can't. The scores should be a bit closer, but I wouldn't expect hm to be identical. My bank's reported FICO score to me is 850, but my Transunion score generated through Credit Karma is 820 and varies 5-10 points from that. I'd look at the reporting dates for her accounts, and pay the amount in full before that date. My main credit card cuts the bill on the 15th, but reports the balance on the 30th. So, on the 29th I pay that full amount, including any charges posted since the bill was created. That keeps my utilization close to zero.

You also said “one of her cards”. I’d experiment, paying that card to zero and have no use at all for a few months. See what impact that has.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.