@Anonymous wrote:So a balance on 8 cards is not more valuable than a balance on 1 card(possibly the opposite) ?
OP, could you list out what your MyFICO scores are today?
Which cards are you letting balances report, and what sort of balance level, relative to each CL, have you been letting report (if any).
@Anonymous wrote:Hi Jamie123. Not sure what a good chunk of your most recent post was about.
You spend a lot of it talking about how the FICO score is only one thing mortgage lenders use when they are deciding whether to lend money to someone to buy a house, and if so, what terms to propose to the potential borrower. I agree. But I haven't at any point suggested anything differently. In fact none of the discussion thread has at any point talked about mortgage lenders and house buying. (I don't think so anyway.)
The OP had a question, I made a pretty standard answer to it, and I added a tiny clause that you and Revelate were a bit doubtful about, so you guys began asking me more about why I thought that.
Looking back, is it possible that you (or Revelate) think I am saying that the FICO model uses raw dollar amount cutoffs instead of % utilization? I definitely have not been saying that. The main factors affecting credit cards (from the Amounts Owed category) that people typically need to think about are % utilization for each account and % utilization overall, plus number of tradelines reporting a positive balance. I then added that if a person wanted to possibly wring a fraction of a point extra, then he might want to consider the raw dollar value owed as well. So I suggested that this was an additional and (I felt) minor factor in the model.
I am happy to keep talking about the general question of whether FICO does indeed look at the actual dollar amount owed, but it would be a shame if the conversation thus far is based on a misunderstanding.
Not trying to speak for Jamie but certainly what I was getting at: I don't think your final assertion has any legitimate basis and it is contrary to not only the likely design goals of FICO but also against all anecdotal reports we've seen here and elsewhere regarding people testing the algorithm.
There may have been a misunderstanding by me but I at least thought you meant "in addition to", rather than "instead of"
The sources you've quoted honestly don't have any credibility as I and others have attempted to point out - FICO pulling out some data and giving it to the public isn't going to be a representative sample of their very closely guarded algorithm in any shape form or fashion. It'd one thing if they were a monopoly, and while they may be close in practical terms, they aren't, and are fighting to retain marketshare. They already catch flak from some sides that the score is manipulatable if you know certain things about the algorithm... and to some degree I agree with that - most people don't pay their balances any time before they have to, and those that are making payments before the statement cuts, are likely statistically "cleaner" with higher scores than the the norm.
End of the day, anyone who wants to figure out what their balances should be, can subscribe to one of the FICO monitoring solutions, and play with various balances on their own credit report. Citing that something is accurate when it's just an estimate or conjecture, is generally frowned on here or elsewhere. I won't go so far as to state there's no way in hell FICO does what you state even on the minor side (though I'm highly suspect of it), but when there's no real source to back it up or even relevant anecdotal data, that's hard to ignore when someone is suggesting it is anything more than rumor... "that's because FICO looks at" rather than something akin to "maybe FICO looks at" in your original response.
We do know the following as you suggest:
- Aggregate revolving utilization counts
- Single tradeline revolving utilization counts (at least maxxed out card, I didn't finish testing that personally and haven't seen others but maybe the precise percentages are know known... I personally think the range might be much higher than what other people have suggested, >9%, but I don't have even good personal data to back that up)
- Number of revolving tradelines reporting a balance compared to number of revolving tradelines aggregate likewise counts
Beyond that revolving utilization, given the vagaries of individuals reports, there isn't really anything that can be stated on specific dollar amounts; however, Jamie is correct in stating that individual dollar amounts likely can't be used as any individual comparison from a FICO perspective, and there isn't enough data available to the model to make a useful prediction from that anyway. Underwriting, absolutely I agree is going to tally up debts vs. assets (along with income) at some level which probably is done in dollar value, but FICO is much more restrained in what it can't and can't do.
@Anonymous wrote:@Anonymous Revelate. Thanks for your response. Toward the end, you wrote:
"... Jamie is correct in stating that individual dollar amounts likely can't be used as any individual comparison from a FICO perspective..."
I wasn't sure what that meant. Is that just a restatement of your belief that dollar amounts owed are not part of the FICO model, only % utilization? Or are you more particularly affirming Jamie's belief that if FICO had done so, they would be exposing themselves to massive lawsuits? Just unclear what you meant.
It sounds like it may be best for me to see if I can find some more sources, beyond the ones I have given you so far, and then perhaps we can discuss further if you like.
Supporting Jamie's opinion though in re-reading I probably could've reworded that to be more clear; lawsuit maybe, but my own assertion is dollar values are pretty irrelevant unless other other data is known, and that data isn't on a credit report.
If I ever get to heavy spending again I'll do a little more testing, but I didn't see a difference in my score at one point from <100 to ~4100 on one of my larger tradelines, score didn't blink which even I was surprised by, but while my own optimum score seems to be >1% it may well be able to go a lot higher than that too. An instrisic dollar amount, doesn't make sense from personal experience and what everyone else has previously posted here and elsewhere for those that have tested the algorithm with varying degrees of rigor.
Not trying to shut down discussion or what not, but it's important to keep theories away from, well "overwhelming anecdotal evidence ~= fact" I guess would be the best characterization. If it's a theory, "hey guys what do you think about this?" is different than "FICO does XYZ in their algorithm." That's where I right or wrong drew the line, probably would've been more coherent and less strident if it had been phrased as a theory to start. Mea culpa.
@Anonymous wrote:One of the places I have heard this discussed is indeed myFICO. So it may be a bit of an overstatement to say that what I mentioned was in contravention of "what everyone else has previously posted here and elsewhere." I will look for the specific link later this week. Hopefully I can find it. Delighted to chat with you further off the thread, if you like. Best wishes...
Public discourse is good for everyone 
.
I've been wrong before, if you / someone else demonstrated it (though I personally think it'd be near impossible to test) absolutely I'm interested in knowing about it. I'm just skeptical .
Not to kill a dead horse...Well...The horse looks pretty dead to me already!
I'm just saying that credit scoring is already a very hot button issue. If FICO were to use dollar values at all, they would open themselves up to lawsuits for discrimination, either from the poor or the rich. (The lawyers don't care which side they are on as long as they get paid!)
It would almost take an Act of Congress to have FICO use dollar values. Remember when a year or two ago Obama wanted medical collections of less than $100 not be used in the score calculations? That didn't go very far.
Just look at the front page of the CFPB's website to see what I mean. The great majority of the page is dedicated to inform people on how to spot discrimination in lending. Ratios don't discriminate but dollar values would. You would see lawsuits if they used dollar values but you don't see lawsuits and hence my reasoning. (I know...It's kinda roundabout!)