How much improvement?

[Anonymous]

This month, I have paid off one of my cc's that was at 100% util, one that was at 98% down to 8%, and paid off a charged off account. My FICO score was at 598 and went up to 602 when I paid off the 100% cc. How much improvement can I expect from paying off the other two accts? The score simulator predicts that it will be at 612-652 but is I add in one month of aging the number jump up to 672-712. Isn't that a big difference? On what date does FICO figure it to have aged a month?

[Anonymous]

Hard to say, but if you're in the 600s, getting utilization down like that is a guaranteed boost to your scores, probably in the 50 to 75 point range.

As far as why that extra month makes a difference, hard to say. I suspect it's a glitch in the simulator, but FICO scoring always has a way of surprising us.

[GFer]

Ooooohhh! I bet you get major points like NWM says! I'm envious!

Great job!

[haulingthescoreup]

---

@Anonymous wrote:
This month, I have paid off one of my cc's that was at 100% util, one that was at 98% down to 8%, and paid off a charged off account. My FICO score was at 598 and went up to 602 when I paid off the 100% cc. How much improvement can I expect from paying off the other two accts? The score simulator predicts that it will be at 612-652 but is I add in one month of aging the number jump up to 672-712. Isn't that a big difference? On what date does FICO figure it to have aged a month?

---

For the one month score hop question, check the opening dates of your accounts, and the dates that any baddies were posted. Are any having a "birthday" next month?

[Anonymous]

You're so smart! I just looked, and our "youngest" account, our car loan, does turn one in August. That must be the projected bump. Will it jump Aug 1, the actual anniv. date, or some combination of the two?

[haulingthescoreup]

You'll get the main bump for the birthday, which is "celebrated" on the first of the month. And then on the actual date of the inq, you'll get a smaller bump for the inq no longer counting, although it will display on your full reports for another year.

That's an impressive projected jump for a birthday! I'm hoping for something similar when my app blitz hits one year starting next February.

[RobertEG]

Expected improvements due to changes in revolving (credit card) utilization....  IMHO

Assumptions:

1. FICO credit util is 30% of total FICO score, per FaurIsaac.  That is 255 pts in the total model (30% of 850)/

2. Of that 255 pts, the large majority is for  revolving util and a small minority goes to installment %util.    So an assumption must be made as to the relative importance of revolving util in the model.  Assume that 240 pts goes to revolving (approx 80%), and 15 pts goes to installment util..  Just an assumption, but needed to generate numbers.

3. FairIsaac has stated in their webinars that their model scores overall util on all revolving accounts about equal to the effect of the util levels on individual cards.  So I then take the 240 ponts, and assign half in scoring to overall % util, and half to the combined effect of individual card util levels.

4.  The final assumption needed to do a mathematical score model is the relationship between %util level, and resulting point score loss.  The simplist mathematical projection is that there is a linear relationship between %util level and point loss. Again, an assumption, but one that seems to track the data I have accumulated, and is necessary for modeling.

OK, with all that in mind, here is what the mathematical modeling shows:

Look first at overall revolving %util  for all cards (120 max, half of total 240 pt impact).

60% util           costs -39 pts

50% util           costs -27 pts

40% util           costs -17 pts

30% util           costs -10 pts

20% util           costs -4 pts

<10% util        diminimus.

Then, the impact of individ card utils must be considered.  If indiv and overall %util count about the same, per FairIsaac’s prouncement, then one need simply divide the above numbers by the number of open accounts to arrive at the impact of each indiv card %util. on score.
So, assuming that one had six open CCs, then each card would additionally impact as follows:

60% util on a single card         costs an additional -6 pts per card at this level

50% util on a single card         costs an additional -5 pts

40% util on a single card         costs an additional -3 pts

30% util on a single card         costs an additional -1.5 pts

20% util on a single card         costs less than one pt.

10% and below                       dimininmus

Finally, from all the posts on the forums, it appears that the FICO algorithm additionally scores the number of cards carrying balances.  With six accounts, for example, and three carrying balances upon report date, then 50% would be balance bearing accounts.

My data shows that this can impact up to -20 pts if all cards carry balances, but again, just my anecdotal observations.  Assuming this to be the case, then the following additional impacts are mathematical:

100% of cards showing balances:       costs -20 points

80% of cards showing baances           costs -16 points

50% of cards showing balances          costs -10 pts

20% of cards showing balances          costs -4 pts

I do not attest to the accuracy of my assumptions or methodology, but submit this is the kind of analysis that one must do to answer the myriad of “what if” questions that permeate the forum.  FICO is math.

These are my estimations of revolving credit util impacts on scoring.