@Anonymous wrote:
Just a data point here in taking aggregate utilization from 1% to 4% due to taking a single card from 0% to 48% (47.5% rounding up) resulted in a 4 point drop on my EX FICO 08 score. Number of cards with a balance reported went from 2 to 3 out of 8, so that factor wouldn't have played a role in the drop.
Hi BBS! Happy new year.
I wouldn't be certain about the text in blue. It assumes a few things:
(1) That FICO doesn't count the raw integer number of accounts -- that it only counts ratios of accounts (e.g. percentage of accounts showing a balance). In this hypothesis, a person with 32 open cards and 8 cards showing a balance would be affected exactly the same as a person with four cards and one showing a balance). If raw integer number is also considered in some models, then 8 accounts might well be scored differently than 1.
(2) That in this ratio (or possibly also raw integer count) only revolving accounts are considered (installment accounts it is being assumed are ignored).
(3) That even if all that matters is a ratio, and the only thing in that ratio that matters are cards (not loans), that there is no penalty as long as the percentage is < 50%. Again, the 50% breakpoint is a hypothesis.
All three are hypotheses. They might be all true, but as far as I know we do not know that any of them are true for certain, and I'd be even less confident of saying that we can be certain of all three simultaneously
Some reasons to have some measure of doubt:
* The relevant reason statements in question (too many accounts with balances) do not specify a ratio -- the language is broad enough to include the possibility that raw number is considered.
* The relevant reason statements do not specify that the accounts are revolving, simply that they are "accounts" -- which leaves open the possibility that installment accounts may be counted in some models.
* I have received these reason statements when I had 2 out of 12 cards reporting a balance. I also had multiple loans reporting a balance at the time. Thomas Thumb (if I remember right) guessed that the penalty was being triggered by a total raw number of accounts (R and I together) being too high. His guess I think was that (as far as the raw number issue goes) 3 or fewer was safe.
Naturally ratios may also matter and I would strongly expect that they do. But I don't think we can be certain that raw number does not. And even if raw number does not matter, there is still the matter of assumptions 2 and 3.
All this is by way of saying that when an extra card began reporting a balance, it's conceivable that this might have been the culprit. Unless the total number of cards (and accounts) showing a balance stays fixed, the experiment is somewhat of a dirty test tube.
