Does momentum count?

[SouthJamaica]

I was wondering if FICO just takes a snapshot of a moment in time and applies its algorithms, or if it factors in momentum or direction.

I.e., if a balance, or total of balances, decreases over time, or increases over time, does the decrease or increase itself factor into the calculation, or is it simply the new numbers that are factored in?

[Revelate]

---

@SouthJamaica wrote:

I was wondering if FICO just takes a snapshot of a moment in time and applies its algorithms, or if it factors in momentum or direction.

I.e., if a balance, or total of balances, decreases over time, or increases over time, does the decrease or increase itself factor into the calculation, or is it simply the new numbers that are factored in?

---

I've only heard one suggestion years ago that a pattern of utilization could bucket you.  Never saw any confirming thing, wasn't much data around it, and that was that.

In terms of pattern of payments, it can't really.  If we get more trended data (balance reported over time) newer algorithms could do something like that, but as of today it's just too inconsistent.  I'm sure it's been looked into and Fannie/Freddie were babbling about new mortgage qualification guidelines for people who just used CC's as transactors, but I don't think it was FICO yet, though I'm sure it's been discussed.

At least all the anecdotal data we've seen, it's instant in time for FICO... but I'd bet money that lenders who keep snapshots over time via their AR process, actually do track it.  FICO doesn't keep snapshots, so no real way for them to do it in the algorithm.

[SouthJamaica]

---

@Revelate wrote:

---

@SouthJamaica wrote:

I was wondering if FICO just takes a snapshot of a moment in time and applies its algorithms, or if it factors in momentum or direction.

I.e., if a balance, or total of balances, decreases over time, or increases over time, does the decrease or increase itself factor into the calculation, or is it simply the new numbers that are factored in?

---

I've only heard one suggestion years ago that a pattern of utilization could bucket you.  Never saw any confirming thing, wasn't much data around it, and that was that.

In terms of pattern of payments, it can't really.  If we get more trended data (balance reported over time) newer algorithms could do something like that, but as of today it's just too inconsistent.  I'm sure it's been looked into and Fannie/Freddie were babbling about new mortgage qualification guidelines for people who just used CC's as transactors, but I don't think it was FICO yet, though I'm sure it's been discussed.

At least all the anecdotal data we've seen, it's instant in time for FICO... but I'd bet money that lenders who keep snapshots over time via their AR process, actually do track it.  FICO doesn't keep snapshots, so no real way for them to do it in the algorithm.

---

Thanks Revelate

[JLK93]

---

@Revelate wrote:

---

@SouthJamaica wrote:

I was wondering if FICO just takes a snapshot of a moment in time and applies its algorithms, or if it factors in momentum or direction.

I.e., if a balance, or total of balances, decreases over time, or increases over time, does the decrease or increase itself factor into the calculation, or is it simply the new numbers that are factored in?

---

I've only heard one suggestion years ago that a pattern of utilization could bucket you.  

---

If we are thinking of the same reported incident, I seem to recall that the person failed to mention that he opened a new account a couple of weeks after the increase in utilization. IMO, his rebucketing, by EQ04, was caused by the new account not the change in utilization.

My experience has always been that FICO is a moment in time.

[Revelate]

---

@JLK93 wrote:

---

@Revelate wrote:

---

@SouthJamaica wrote:

I was wondering if FICO just takes a snapshot of a moment in time and applies its algorithms, or if it factors in momentum or direction.

I.e., if a balance, or total of balances, decreases over time, or increases over time, does the decrease or increase itself factor into the calculation, or is it simply the new numbers that are factored in?

---

I've only heard one suggestion years ago that a pattern of utilization could bucket you.  

---

If we are thinking of the same reported incident, I seem to recall that the person failed to mention that he opened a new account a couple of weeks after the increase in utilization. IMO, his rebucketing, by EQ04, was cause by the new account not the change in utilization.

My experience has always been that FICO is a moment in time.

---

It is entirely possible I didnt keep full track but I was severely skeptical when I read it.

[Anonymous]

This thread reminds me of one about 3 months ago that I started in posing the question of whether or not utilization should have some sort of memory:

http://ficoforums.myfico.com/t5/Understanding-FICO-Scoring/Utilization-has-no-memory-But-should-it/t...

I know you were part of that thread SJ, although I don't remember you exact opinions.

I liked how the discussion went into trended data and considering a bunch of other hypotheticals that could be used in scoring (but aren't) such as monthly spend, interest paid, payment patterns, utilization history, average daily balance, etc.

In short I think there are a lof of other metrics out there that would provide far more useful scoring data outside of the simple moment in time snapshot that we currently get with the utilization sector of scoring in particular.

[Anonymous]

To reaffirm what the other folks are saying, no the existing and widely used FICO models base everything on a snapshot.  Patterns of utilization, or of the amount of payments against reported statement balances, or anything like that, are not considered.

But there seems to me no doubt that future FICO models (e.g. FICO 10 and likely Vantage 4.0) will use those kinds of trended data, since they are such powerful predictors of risk.  (And they are much less easily "gamed" than the traditional "pay off your credit cards a month before you buy your house" technique)  FICO didn't use TD in the past because all of the widely used FICO models were designed at a time when these trended data did not exist.  The CRAs were simply not collecting them.

I think it is also quite likely that FICO and other groups are right now creating specialty products to sell to lenders who wish to supplement their decision making with a TD analysis of your report.  Fannie Mae included TD analysis software inside the latest version of its Desktop Underwriter software -- if I remember right the TD-analysis module was designed by some folks at TransUnion.  And any number of credit card issuers may be using their own proprietary systems that integrate a mainline FICO model with other in-house concerns.

So certainly there's no reason for consumers who now carry balances not to be transitioning toward a PIF approach, since they will be creating a Transactor history that will be useful down the road (if not now).  Being a transactor always made financial sense.  This is just one more reason to do it.

[SouthJamaica]

---

@Anonymous wrote:

This thread reminds me of one about 3 months ago that I started in posing the question of whether or not utilization should have some sort of memory:

http://ficoforums.myfico.com/t5/Understanding-FICO-Scoring/Utilization-has-no-memory-But-should-it/td-p/4738505/highlight/true

I know you were part of that thread SJ, although I don't remember you exact opinions.

I liked how the discussion went into trended data and considering a bunch of other hypotheticals that could be used in scoring (but aren't) such as monthly spend, interest paid, payment patterns, utilization history, average daily balance, etc.

In short I think there are a lof of other metrics out there that would provide far more useful scoring data outside of the simple moment in time snapshot that we currently get with the utilization sector of scoring in particular.

---

Well to be perfectly honest with you, I wasn't looking at it in such a scientific, high minded, way, wrestling with how it should be and what would be philosophically the best way to gauge creditworthiness.

I was just wondering, if I purposely let a higher balance report, and then paid it down gradually, would there be any points in that

[Revelate]

Only if you were trying to show a lender something; it might work to your advantage to be "carrying a balance" but I would suggest that's not really needed.

Until I start getting denied or others do for being a transactor I'm sticking with it.

Can fakesy though on a 0% offer but meh.

[Anonymous]

---

@SouthJamaica wrote:

---

I was just wondering, if I purposely let a higher balance report, and then paid it down gradually, would there be any points in that

---

There would be no extra FICO points in it for you, and your initial post asked about FICO.  (Presumably you meant FICO 8 or any of the other models currently being used by a significant number of lenders.)

Down the road it would if anything have been harmful to you, since you would have created history for yourself as a revolver.  It's conceivable of course that eventually specialized software will be sophisticated enough down the road to identify people who are revolvers but also are utra low risk.  That would be a nice group for CCCs to target.  But if it ever happens that CCCs stop issuing you credit cards because you don't ever carry balances, you can change your behavior then.  Of course, by then, at your current rate of CC acquisition, you will probably have 80 credit cards.  :-)