Score Increase Over Time

[Anonymous]

Depending on the CRA, 20-30 points from that.  Unfortunately it's hard to get an exact number because not only did that 95% utilization post, but because of Irma I was "trapped" on a small boat in the gulf for 4 days and was unable to PIF my new charges on my Discover which reports 42% on that (nornally I keep that card at $0 every statement).  So instead of just one card reporting a 95% utilization ($490/$12600 = 3.8% aggregate utilization) I actually have 3 cards reporting a balance ($5 on my AZEO card, $490 on Capital One, and $1070 on Discover) which has me reporting $1565/$12600 or 12.4% aggregate so I also crossed the 9% aggregate threshold.  Annoying but not unexpected considering I wasn't able to get online at all during Irma unless I wanted to pay $50 in satphone data charges, lol.

3 cards reporting a balance definitely appears to affect Equifax FICO the most but I feel confident that it affects all of them in some way.

So I'm hoping in a week or three to see what recovers.  Will update once everything new reports.

I know Discover will manually force an update midcycle if requested but it isn't a huge deal right now because I'm not credit-seeking at all.  I already PIF'd Discover to zero upon making landfall and picking up a mobile signal, so I doubt that will preclude any graduation when my 7th statement posts in November.  I also assume Capital One will report $0 as the account has been $0 since the day after the unexpected statement cut.

I'm pretty consistently on keeping all my balances to $0 (other than AZEO) since I push a payment as soon as I charge anything significant, otherwise I just push payments weekly to keep things at $0 (or a negative balance, even).  Just easy to do from Chase on mobile, takes all of 15 seconds.

[Thomas_Thumb]

---

@Anonymous wrote:

Having cards report high individual utilizations can put many people at risk of adverse action by creditors.  Note that this is not just the creditor for the card you are maxxing out but also other other creditors, since they can see your reports and scores each month as well. 

I can't count the number of aggrieved people who write in to the forums demanding an explanation of why Citi closed their card when they "had never missed a payment" on the Citi card -- though by the second or third post it becomes clear that he had one or more cards from other issuers maxxed out.

Some factors sharply reduce this risk, such as having a low overall utilization, a clean profile (no derogs), and some significant payment history; and a FICO above 720.  A person for whom all those things are true can likely allow individual cards report high utilizations (even for extend periods of time) without risk of AA.

As far as trended data, we simply can't say for sure today how it will play out.  Some people have reported that many CC issuers stopped reporting their history of monthly payments in 2016.  If that is true, and it remains true in 2018 and onwards, then it won't be possible for a creditor using TD to detect whether a person pays his CC statements in full each month.  (Such people are called Transactors.)  But if we do move into a world where detecting Transactors is possible, there's no question that loan issuers will care about that a lot.  It's been proven that people who do not PIF are statistically (as a group) far riskier to the lender than those who do not.  That's why Fannie Mae added a module to its Desktop Underwriter tool that attempts to detect whether a prospective borrower is a transactor.

I plan to pull all three of my credit reports at ACR early next year to see which of them are reporting the month-by-month history of all my payments. 

---

I can confirm that payments have not been listed on my CRA reports since mid/late 2015. However, that does not mean such data is not reported to CRAs.

I obtained a full 3B report through PrivacyGuard as part of a trial back in April 2017. Interestingly, two of theaccounts on my PrivacyGuard report showed payment data which I had not seen on the CRA reports pulled in March 2017. So my conclusion to equate lack of data being displayed on CRA reports to lack of data being reported appears flawed. 

CGID, If you have CRA reports for all three bureaus from 2016 compare the 2015 and 2014 account payment field data by CRA. My recollection is two CRAs listed payments (TU being one) and the 3rd did not. If a CC account is reported to all three CRAs, lack of specific data being displayed by one CRA does not equate to lack of availability. CRA reports typically only show monthly account detail for the previous 2 years. So your 2018 reports would not include payment details for 2015 and 2014.

[Thomas_Thumb]

---

@Anonymous wrote:

---

@Thomas_Thumb wrote:

---

---

IMO - useful trended data analysis needs to segment according to transactor/revolver behavior as a 1st step. If one is a transactor reported balance/utilization is a non factor and should be ignored - like is done on a true charge card. From a risk perspective, a PIF before statement cuts is no different than a PIF after statement cuts given the same monthly charges in either case. Reported utilization is more meaningful for as a risk factor for revolvers.

I do agree that total monthly charges (either in absolute terms or as a % of CL) can help detect increased risk levels if trended even for transactors. As an example, a month over month increase in total spend (or aggregate utilization %) over the past 6 months suggests someone may be approaching a spend level that cannot be maintained before a subsequent late payment or default occurs. This metric would be most useful if it were to be based on total charges during a month - not just what is allowed to report on a statement as a balance.

---

So TT, in my above example of the person reporting 90% utilization on a $500 limit card month after month where the first person is making a $450/mo payment (and running the limit back up over the next cycle) and the second person is simply leaving that balance on there, just making the minimum payment, wouldn't trended data result in these behaviors being viewed very differently?  I mean, under a MR wouldn't the person making the large monthly payments be seen as far less of a risk than the person carrying the balance, even though it's a toy limit card?  This is a situation where I feel the utilization percentage, not the dollar amount, plays a significant role.

---

Yes and no. Trended data looks for changes in patterns/behavior for early detection of a change in risk level. In both cases trended data would indicate consistent behavior - equating to no increase in behavior/risk month to month. Both are "managing credit".  Now if transactor/revolver analysis were added to the algorithm, the 2nd person would be considered a higher risk than the 1st. Under a manual review, I would agree with you that the 1st example demonstrates less risk - because it demonstrates the person has the means to fully pay established debt.

Use of trended data is kind of like SPC. 

1) is the process (behavior) in control?

2) Are there multiple consecutive data points showing an upward or downward trend (say 7)?

3) Are two of the last 3 data points more than one standard deviation above/below the mean?

4) Is the last data point more than 2 standard deviations above the mean?

 * Items 3 & 4 require many months of data to establish standard deviations. Of course pre-defined thresholds could be used in place of standard deviations - credit monitoring is not a controlled process after all.

[iv]

---

@Anonymous wrote:

As far as trended data, we simply can't say for sure today how it will play out.  Some people have reported that many CC issuers stopped reporting their history of monthly payments in 2016.  If that is true, and it remains true in 2018 and onwards, then it won't be possible for a creditor using TD to detect whether a person pays his CC statements in full each month.  (Such people are called Transactors.)  But if we do move into a world where detecting Transactors is possible, there's no question that loan issuers will care about that a lot.  It's been proven that people who do not PIF are statistically (as a group) far riskier to the lender than those who do not.  That's why Fannie Mae added a module to its Desktop Underwriter tool that attempts to detect whether a prospective borrower is a transactor.

I plan to pull all three of my credit reports at ACR early next year to see which of them are reporting the month-by-month history of all my payments. 

---

Current trended-data availability that I'm aware of as of this month:

Amex: only reporting account balance

BoA: reporting account balance, minimum payment amount, and actual payment date

Chase: reporting account balance, minimum payment amount, and actual payment date

Citi: reporting account balance, minimum payment amount, and actual payment date

Discover: reporting account balance, minimum payment amount, and actual payment date

Synchrony/Amazon: Full data, including actual payment amount (started full data Feb 2016)

Wells Fargo: Full data, including actual payment amount (has been full data "forever")

All also report credit limit history.

This data is now appearing consistently across all three CRAs, for the lenders that are actually reporting.

As previously discussed, some of the lenders that are currently NOT reporting full data have done so in the past, and stopped.

(Unlike revolving lines, where this data is hit-or-miss depending on lender, installments do seem to consistently populate these fields... although that data isn't nearly as useful for lenders to analyse.)

[NRB525]

---

@Anonymous wrote:

Hello everyone, in recent months I've gotten my first card, the Discover Secured, as well as becoming an AU on various family members' cards and getting really interested in keeping a high credit score. Since I won't have my first official FICO until December, I've had too much time reading up on the best ways to maximize your score and the importance of utilization and so on in these forums.

My question is that since utilization is only a snapshot and has no memory effect, what are the main drivers in causing a score to rise over time? Assuming that a person keeps the ideal utilization and number of cards for months, is the only FICO factor that changes and therefore causes the score increase the age of accounts? Since it only accounts for 15% of the score overall that would be pretty surprising but also interesting to note. Some have shown that initial scores ~700 or less is pretty common and I was just wondering what would cause it to increase to the higher 800 range from there if utilization and payment history were already ideal. I'm just really anxious to get my first real credit score and start building from there, and I've just been obsessing over the tiniest of details that won't matter for now and making sure that I get off to the right start with credit.

Thanks for your insights.

---

Getting back to the original topic, which got an initial answer, but then branched wildly...

The main thing about starting out with a score is, there is no history. The main point of FICO scoring is to try to find one number that gives lenders some indication of risk of default by this borrower.

When you start out, you get something in the range of a 700 score, as a gimme. From there, it is up to the cardholder to show that they can make payments on time. That's really all there is to it. If you make every payment on time, your score will increase.

If you miss a payment (or a bunch of payments) then your score is reduced, because your risk is higher. "Missing payments" covers a whole range of things lenders would rather not see, including bankruptcy. Credit is all about trust, and this missing payments thing is a trust killer.

Beyond that, the discussion around Utilization becomes a factor influencing your score. Utilization is a modification factor, after your payment history is setting the baseline. AZEO puts you at the top of the range of possibility, based on your payment history. AZEO by itself cannot get you an 850 score. Higher utilization, such as I had years ago, holds down your score, but not to the same degree, and not for as long a time, as missing payments. High utilization is a risk factor for default, on the basic assumption that a lender only feels comfortable extending X amount of credit. If you use 90% of X, that is considered a Risky Thing. A high percentage of accounts reporting, and significant amounts outstanding, like I still have, holds a score down. Adverse Action by lenders who do not like the risk level is a factor in raising utilization, lowering score. But AA is a reflection of risk factors that lenders do not like, so if AA is happening, the borrower needs to look clearly at why, and determine if borrower action is necessary.

In answer to your fundamental question, however, only the passage of time will get your score increased. Months and years of on time payments builds up that trust, shows you are trustworthy, and raises your FICO score. Utilization may bring it down, number of active cards may bring it down, but utilization is a choice of how to use credit. If I were able to do AZEO, I'm sure I would have an 850 score. But then I would not be borrowing any funds.

[Anonymous]

Lots of interesting info about trended data discussed, and I had one more minor question to throw into the mix.

Since utilization percent has no memory (for FICO scores), is the same true for number of accounts that report a balance? Regarding things such as AZEO. If trended data were to be used to judge creditworthiness, would varying from each month the number of cards with a balance negatively impact you like it slightly does to your score when more than half show a balance? If utilization were still kept low but say 3 out of 5 cards show a balance instead of 2 or 1 etc.

Just regarding my personal situation where I’m an AU on three cards and it’s not practical to have them all show zero though we do pay it down before the statement cuts. I’m guessing that most creditors won’t care too much the number of accounts with a balance as long as utilization is kept low and no late payments of course. But then again that’s why I come here for answers 😄.

[Anonymous]

---

@Anonymous wrote:
Lots of interesting info about trended data discussed, and I had one more minor question to throw into the mix.

Since utilization percent has no memory (for FICO scores), is the same true for number of accounts that report a balance? Regarding things such as AZEO. If trended data were to be used to judge creditworthiness, would varying from each month the number of cards with a balance negatively impact you like it slightly does to your score when more than half show a balance? If utilization were still kept low but say 3 out of 5 cards show a balance instead of 2 or 1 etc.

Just regarding my personal situation where I’m an AU on three cards and it’s not practical to have them all show zero though we do pay it down before the statement cuts. I’m guessing that most creditors won’t care too much the number of accounts with a balance as long as utilization is kept low and no late payments of course. But then again that’s why I come here for answers 😄.

---

You may need to clarify your question.  Are you asking how current and past FICO models work?  The text I highlighted in red suggests this is what you want.  But in the text in green you switch to asking how models that use trended data would work.

The crucial thing to bear in mind is that utilization has no memory in FICO models precisely because these models do not use trended data.  Therefore other factors like number of accounts with a balance also have no memory in current and past FICO models.  "Memory" is by definition a function of trended data (with the exception of derogs, which have had a 7-year backward memory since the 70s or even earlier).

The best practical plan (as I mentioned earlier in this thread) is therefore this:

Use your cards as needed.  Let them report positive balances (if you happen to use the card).  Always pay the statement balance in full.  But keep your utilization under reasonable control.

Then use AZEO in the month before you want to squeeze out every extra conceivable point, i.e. typically before you apply for a loan or a new card.

That's a practical plan because in practice most people do not need their score to be optimized every single month.

[Anonymous]

Sorry I was unclear. I meant that since all parts of utilization including number of accounts with a balance have no memory, and it has been shown that not following AZEO gives a very slight drop in FICO score at the time a score is generated, would creditors also take this as a negative mark if they saw in trended data that a “non-ideal” number of accounts showed a balance. Even if they all had low utilization and are still dinged under the current scoring model. So broadly, is it likely that a creditor would also view this as a negative for having multiple accounts with low utilization, like is indicated by the scoring model?

I realize this is getting to the minute details that may not even apply in the future, but I was just curious after reading the replies in this post. I see the practical plan going forward you outlined and will probably follow it anyways, but I’m always wondering what the “perfect” behavior is in the eyes of the creditor if that makes sense haha.

[Anonymous]

Like utilization, there are varying degrees of adverse payment history.  Payment history and utilization carry similar weight in FICO score impact, 35% vs 30%.  Both of these factors need to be in a good place to posses top FICO scores; You can't ace one and bomb the other.

While utilization does make up a slightly smaller slice of the FICO pie, if you're talking maxed-out utilization it can often destroy a FICO score worse than missing some payments.  Payment history typically can't go from great to horrible in 1 moment, as payments are reported late in steps:  30 days, 60 days, etc.  One doesn't just have 120+ day late payments appear on their credit report, at least that's not how it usually works.  That being said, if maxed-out utilization is holding back a score say 150 points, the first late payment someone receives (30 day) usually results in a ding much less than that.  From what I've seen, most people report a loss in the 70-100 point range.  Now, if you're talking major delinquencies, that's a different story.  It is important to realize though that utilization can be fixed with money (paying off that debt) and recovered from as quickly as one is able to do so, where payment history can really only be fixed with time (say 7 years).  There are people out there that are able to achieve the removal of negative payment history items, but relative to the whole they are extreme outliers, so I'm not considering them for this discussion.  My point here is simply that in terms of FICO score, maxed out utilization can be more damaging than minor payment history hiccups, especially in the short term.

[Anonymous]

---

@Anonymous wrote:
Sorry I was unclear. I meant that since all parts of utilization including number of accounts with a balance have no memory, and it has been shown that not following AZEO gives a very slight drop in FICO score at the time a score is generated, would creditors also take this as a negative mark if they saw in trended data that a “non-ideal” number of accounts showed a balance. Even if they all had low utilization and are still dinged under the current scoring model. So broadly, is it likely that a creditor would also view this as a negative for having multiple accounts with low utilization, like is indicated by the scoring model?

I realize this is getting to the minute details that may not even apply in the future, but I was just curious after reading the replies in this post. I see the practical plan going forward you outlined and will probably follow it anyways, but I’m always wondering what the “perfect” behavior is in the eyes of the creditor if that makes sense haha.

---

We simply don't know how creditors or future scoring models will use trended data.  At one time it seemed certain that all CC issuers would be reporting the amount of that a cardholder paid each month, but it seems quite possible that many CC issuers are not doing that.  If true, then TD is really weakened as a predictive tool.

A reasonable thing to do is to keep a foot in both the old pre-TD system and in the TD approach of the future (if it ever materializes).  So know how to maximize your score via the old approach (AZEO the month before an important pull) and for the future always make sure you are paying your cards in full.  The latter is the defining mark of a transactor, which TD was supposed to be able to enable creditors to detect.  Transactors are much less risky than those who carry balances, so a TD-based system would no doubt look at that. 

PS.  The temporary score drop for not using AZEO can be be quite significant (not very slight) depending on the model.  The old mortgage models care a lot about number of accounts showing balances.