10 Point EX Drop for Paying Off Car Loan

[SouthJamaica]

---

@Anonymous wrote:

---

@SouthJamaica wrote:

---

You can call me 'silly' if you like; I wouldn't report you

---

LOL, thanks, I appreciate that

I tend to use the term silly when something is irrelevant to the whole or vast majority.  It's like if 99 out of 100 people are denied for Credit Card X with a score of 680 or less, but 1 person chimes in saying they were approved with a 640 score.  If that person presents an argument suggesting that because they were approved with a 640 score that others will be approved as well, to me while it's factual that they were approved under a 680 score it would be a silly argument to make a blanket statement that people with a 640 score can/will be approved.  Perhaps that example above is silly in and of itself, but hopefully my point comes across.

---

Yes, well that example would be a logically flawed argument.

But these are examples from which it could be reasonably concluded that the FICO 8 algorithm does favor people who are currently in debt over people who are debt free:

Assume A has 5 revolving accounts each with a limit of 10,000, and 1 installment loan with an original amount of 25,000.

Scenario I : A owes 2500 on each of 2 revolving accounts, and is struggling to make the minimum payments each month; additionally he owes $2250 on his installment loan, and is struggling to make a $200 month payment on that.

Scenario 2 : A owes 2500 on each of 2 revolving accounts, and is struggling to make the minimum payments each month, but has just managed to pay off his installment loan.

Scenario 3: A has paid off all his credit cards and is showing zero balances, but continues to struggle with his $200 month loan balance.

Scenario 4: A has paid off all of his credit cards and is showing zero balances, and has paid off his loan, and owes nobody anything and is not struggling to pay anything any more.

As you well know, under Scenario 4 -- where A is in real life the least risky of the 4 borrowers -- his FICO 8 scores will be the LOWEST, and under Scenario 1 -- where A is in real life the most risky of the 4 borrowers -- his FICO 8 scores will be the HIGHEST. And of course under Scenarios 2 and 3, he will be somewhere in between.

These examples are not silly.

Now there are one of two things that are silly, and I can't say with certainty which it is, but it is definitely one of them.

Either

(A) the algorithm is meant to measure risk, and it is therefore the algorithm that is silly, or

(B) the algorithm is not silly because it is really intended to measure not only risk, but also reward, and the only thing that is silly is the argument that owing some money is more evidence of credit worthiness than having paid off one's debts.

I leave it to you to decide which is silly.

[Anonymous]

I see what you are saying and don't think it's a bad argument.  You're not wrong.

But, for the sake of argument, here's my take.  "Struggling" is an irrelevant term for two reasons.  One, Fico scoring (or anything really) can't measure "struggling."  Two, whether someone struggles to make a payment doesn't matter.  It all comes down to if they can make minimum payments or not.  Whether it's easy or a struggle isn't (and shouldn't) be considered.

The algorithm is designed to measure risk.  I don't necessarily think it's necessarily silly, just perhaps behind the curve when it comes to certain things like the couple of quirks we discussed and that you referenced in the various examples above.  Perhaps in a later FICO scoring model for example we'll see trended data used, where a reported balance would become irrelevant and a mere spend on a CC would be seen/considered and constitute revolving credit use, even with [all] $0 reported balances.   

Logicially flawed arguments if they are outlier arguments shouldn't be presented as blanket statements.  I'll give one example unique to only myself as far as I know.  It's well documented on this forum that those that start off in a Capital One "starter bucket" will likely never see a significant credit line increase.  Many people report starting with (say) $500-$3000 limits and seemingly can't get more than a $500 CLI or two and are destined to remain at a low limit forever.  I was able to "cheat" the system after discovering I resided in a starter bucket, having been approved with a 619 score and securing nothing but a pair of $500 CLIs over a length of time.  I contacted the EO and told them that their online and phone (automated) systems weren't working for me and that I needed them to do a MR CLI.  After some arm-twisting, they agreed to and ultimately approved me a CLI to $10,000.  I can say with 100% certainty that if it weren't for the MR CLI, my limit on the starter account would still be stuck at $4000.  Instead I've got a decent 5-figure limit on the starter account that I was first approved with at the start of my rebuild when I had zero revolvers to my name.

Now, if I were to start a thread blanket-statement style suggesting that basically anyone with a bucketed "starter" account could bump their limit up to 5-figures, no bubbles no troubles, it would be quite misleading.  Even after starting a thread about this a couple of years back, to my knowledge not 1 person outside of myself was able to replicate what I did with a MR CLI through CO.  While what I stated was in fact factual, it was without question an outlier example and IMO should never be presented as a blanket statement suggestive that it's easy or the norm.

[Anonymous]

Whether or not we think the algorithm is silly is just our opinion. Whether or not the designers’ intent is met in the best possible way, again is our opinion. Nevertheless I still firmly believe the purpose of the “no revolving balance penalty” as we call it, is to give points to those who are actively using credit. One cannot demonstrate payment history if they are not using actively using credit and therefore can’t be evaluated for creditworthiness. How can you evaluate somebody’s risk of repayment if they are not currently using credit? If you’re using six-month-old data, then your risk assessment is six months old. Currently if you have no revolving balance the algorithm doesn’t take into account how long it has been since you’ve used credit. Maybe that’s a flaw that needs to be addressed?

So I don’t think it’s intended as a penalty but as a bonus for those who are showing active use of revolving credit. Of course we look at it as a penalty but that doesn’t mean it was the intent.

We have break points from 8.9 to 88.9%. But those are totally useless to evaluate your creditworthiness, if you’re not using credit. Therefore, if you are not using it and cannot be evaluated you have a lower score because it’s a higher risk since you cannot be effectively evaluated by the utilization portion of the algorithm. On the other hand, if you can be evaluated, then here are the starter points for being able to be evaluated at the lower end of the risk spectrum. Then as the the risk increases, the the algorithm takes points away as the risk continues to go up.

I’m not saying this is the best way, i’m not saying I like it or support it, and Im not saying it necessarily works, I’m just saying I think this was the intent of the designers.

Said another way you’re not penalized if you don’t use credit but if you do use credit you’re given a certain amount of points and then as your percentage of risk goes up, those points are taken back and even more points are taken at higher risk levels. Does this mean is the best way of evaluating? Again that’s for the designers and statisticians to determine. That’s why they’re continually revising the model, it is not perfect but I believe this was the objectives, whether they were met or not is subject to debate.

And yes struggling is not something that can be measured by the algorithm which makes that an inapposite analogy.

[OmarR]

---

@Anonymous wrote:
Nevertheless I still firmly believe the purpose of the “no revolving balance penalty” as we call it, is to give points to those who are actively using credit. One cannot demonstrate payment history if they are not using actively using credit and therefore can’t be evaluated for creditworthiness. How can you evaluate somebody’s risk of repayment if they are not currently using credit? If you’re using six-month-old data, then your risk assessment is six months old. Currently if you have no revolving balance the algorithm doesn’t take into account how long it has been since you’ve used credit. Maybe that’s a flaw that needs to be addressed?

---

Your last 2 sentences (that I quoted), in my opinion, nailed it.

FICO can take into account negative info for 7+ years. Judgements can be 10 years, even renewed for a total of 20 years. Positive accounts up to 10 years. But they change the game for installment loans.

IMO, installment loans, open or closed, should be 10 years.

[Anonymous]

---

@OmarR wrote:

---

IMO, installment loans, open or closed, should be 10 years.

---

In terms of credit mix, they are.  In terms of the portion of Amounts Owed related to them, that's only considered with an open loan.

[Anonymous]

When I have paid of an auto loan generally saw at least the loss of 20 points ... GRR

[OmarR]

---

@Anonymous wrote:

---

@OmarR wrote:

---

IMO, installment loans, open or closed, should be 10 years.

---

In terms of credit mix, they are.  In terms of the portion of Amounts Owed related to them, that's only considered with an open loan.

---

Which is why i wrote "open or closed". You seemed to have missed that part.

[Anonymous]

As many have observed, there are flaws with the algorithm. And I do tend to agree that it should consider amounts owed longer than immediately. Obviously, if you’ve paid a loan for years and paid it off, I think that should be good to establish ability for some time period, as SJ mentioned. I think the trended data is meant to address this fallacy. We’ll see if and when it comes into effect.

Like Dinosaur said, he suffered a loss when he paid off the loan. I don’t think it should be that way. I think you shouldn’t lose points for at least for a certain period of time. There should be a time period to buffer it, maybe even break it into losing a few 6 months off and the rest a year off, but that’s just my opinion.

[Anonymous]

---

@OmarR wrote:

---

@Anonymous wrote:

---

@OmarR wrote:

---

IMO, installment loans, open or closed, should be 10 years.

---

In terms of credit mix, they are.  In terms of the portion of Amounts Owed related to them, that's only considered with an open loan.

---

Which is why i wrote "open or closed". You seemed to have missed that part.

---

I saw it, but you didn't make mention if you were referring to credit mix, amounts owed, or both. 

[OmarR]

I did mention it, by quoting Birdman’s eloquent post.