Utilization %s

[Logical]

smallfry wrote:

I know in my case the reason the scores went down during that one month period was because I let 3 of 6 open revolving lines report a balance.

That sounds Logical

---

 I imagine if I live to be 100 I could see 800. I started playing the game too late.

---

           When you're 100 you won't need it anymore!

[RobertEG]

The existence of different "buckets" based on peer groupings is not anecdotal.  It is clearly stated as a fact by FairIsaac. They acknoweldge approx. ten peer categories, and state that FICO08 will expand this to around 12, breaking down the lower buckets with more refinement.  The only real question is whether, within each bucket, or peer category, the change in score is refined to the level of looking at risk/raqte of change effect at each instant level of, for example, %util, or whether the risk potential is chunked into tiers, and only changes when the tier level within each category changes at some trigger point. 

The evolving terms of "buckets" and "tiers" are quite different, so I will use that distinction.  Buckets are overall peer groups used to determine which general algorithm will be used.  Tiers are how they choose to construct the model within each bucket algorithm.  Buckets are real.  Tiers are speculative, for FairIsaac is not about to release the proprietary level of detail as to how each of their specific algorithms are constructed, for obvious reasons.

[Logical]

---

RobertEG wrote:

Buckets are real.  Tiers are speculative

---

100% agreement!

---

FairIsaac is not about to release the proprietary level of detail as to how each of their specific algorithms are constructed, for obvious reasons.

---

I think it would also be injurious for us to know it. Except for those (for instance) who need to get a mortgage in the next year, most of the obsession with Fico (will this or that card help Fico the most) is a detriment to financial health. I will bow to Fico and keep my oldest CC, but I prefer to make decisions on what is best financially, not what is best for Fico. (I know that is heretical!) 

[MidnightVoice]

---

@Logical wrote:

---

@RobertEG wrote:

Buckets are real.  Tiers are speculative

---

100% agreement!

---

FairIsaac is not about to release the proprietary level of detail as to how each of their specific algorithms are constructed, for obvious reasons.

---

I think it would also be injurious for us to know it. Except for those (for instance) who need to get a mortgage in the next year, most of the obsession with Fico (will this or that card help Fico the most) is a detriment to financial health. I will bow to Fico and keep my oldest CC, but I prefer to make decisions on what is best financially, not what is best for Fico. (I know that is heretical!) 

---

A lot of us differentiate between Fico sense and Financial sense.

And we go with the latter unless we are looking for credit in the near future.

And when we get most of our problems sorted out and high enough scores and TLs, the difference becomes minimal

[KingAdrock]

Why is there all this talk of good FICO sense and good Financial sense being mutually exclusive? Almost everything you do that's good for your FICO score (paying on time, PIF, keeping balances low) are also good for your financial health.

IMHO, situations where FICO and Financial sense conflict are pretty rare, and even when they do (such as your oldest CC has an AF) Financial sense wins out, even among us "FICO obsessed" types.

[haulingthescoreup]

Probably the best example of a conflict between the two is a 0% APR or BT offer. It makes sense financially to load as much debt onto that as you can, so that you can pay principal only. But it's terrible for your scores, and you risk AA (adverse action) by other CCC's, who see the high util and don't know what you're doing.

Another is the early pay-off on installment loans. No point in paying interest if you can pay it off now, but if it's your only installment loan (or mortgage), paying it off will remove that element from your credit mix, and that often drops your scores.

This doesn't mean don't do it; it just means be aware of what might happen.

[RobertEG]

Excellent points, Hauling, and there are many, many more.  Disregarding FICO totally, one would just blindly apply for any and all new credit, maybe a dozen new apps a month, seeking the best terms and rates for each.  Financially smart, but FICO suicide.  One must have two plans... a financial plan, and a FICO plan.  Which one follows in the short term depends upon one's current situation.   They can be major in financial actions taken.

[RobertEG]

Logical, once again I agree with you (provided you are speaking of tiers, and not buckets).  There is NO doubt that a continuum algorithm, which evaluates risk at every specific percentage level using differential calculus, as opposed to a tier bracketed algorithm, which sets linear tiers for varying the evaluation at different break points, would be more accurate and pecise mathematically. 

So the issue is not whether it is better.  I am sure that the nerd programmers at FairIsaac would agree with you.  But the nerd programmers dont make the business decisions, and apparently FairIsaac has chosen to smooth it out rather than shoot for daily mathematical accuracy. 

It is something we could argue all day long as to which is better.    FairIsaac produces and sells the algorithm, and it appears that tiers are their chosen businees product.   So we live with it, and just try to udnerstand it rather than second guess it.

[Logical]

---

MidnightVoice wrote:


A lot of us differentiate between Fico sense and Financial sense.

And we go with the latter unless we are looking for credit in the near future.

And when we get most of our problems sorted out and high enough scores and TLs, the difference becomes minimal

---

Very true, and those who follow the above will be among the successes 5 years from now.

[Logical]

---

RobertEG wrote:

Logical, once again I agree with you (provided you are speaking of tiers, and not buckets).  There is NO doubt that a continuum algorithm, which evaluates risk at every specific percentage level using differential calculus, as opposed to a tier bracketed algorithm, which sets linear tiers for varying the evaluation at different break points, would be more accurate and pecise mathematically. 

So the issue is not whether it is better.  I am sure that the nerd programmers at FairIsaac would agree with you.  But the nerd programmers dont make the business decisions, and apparently FairIsaac has chosen to smooth it out rather than shoot for daily mathematical accuracy. 

It is something we could argue all day long as to which is better.    FairIsaac produces and sells the algorithm, and it appears that tiers are their chosen businees product.   So we live with it, and just try to udnerstand it rather than second guess it.

Oh, we've always been in basic agreement, Robert. I would just add that the suits are selling accuracy.

I would point out my anecdotal evidence. A month ago a line of credit went from 100% util to 0% util. Scores went up several points. Even if the entire change could be attributed to this, a change of only 10% would have been miniscule. Utilization is only a part of the "how much owed" (app 30%) category. Plus folks with few trade lines would be affected more than folks with a large number of TLs. It does get complex.