@Thomas_Thumb wrote:
@Revelate wrote:Yuppie Foodstamp = $20 bill, so named from yuppies used to pulling out plenty of $20's from the ATM and using them copiously in restaurants late 80's early 90's.FICO doesn't appear to make a distinction between different types of installment loans other than maybe Nextgen, but it's up for debate on FICO 9 potentially.Agreed on FICO 8 = FICO 04 = an inquiry is an inquiry and remains as such for the entire year it's on there. Also agreed that my data appears to suggest that inquiries in the first 3 or possibly 6 months (where's the 5 estimate from?) but we should get close to confirming that in another 29ish days when the mortgage coded inquiry counts against me.
Oilcan12's inquieies at the time of his F09/F08 comparison were 5 months and older - His F08 scores were lower across the board than F09 scores.
That data suggests any theoretical short term impact associated with F09 had been accounted for by month 5. A shorter timeframe than 5 months (90 days or 3 months) is a more logical timeframe but his data did not include anything below 5 months.
We cannot compare FICO 8 and FICO 9 scores to pull data from like that; they weight things differently, and as such direct comparisons on absolute score values such as this are not worth anything unfortunately.
Case in point, my EX 9 score is higher than my EX 8 score, but clearly my inquiry is not at the 5 month threshold and is only at 2 months; therefore we cannot discount 6 months for example as being the breakpoint as yet and certainly can't state that 3 months is the breakpoint either. End of the day we're going to have to compare FICO 9 with FICO 9 when we're talking specific weights to specific things like inquiries. I'm making the not quite blithe assumption that my EX and EQ scores should absolutely be in the same ballpark for FICO 9 as they are for FICO 8 given identical data should be on identical scorecards between bureaus, and given that the obvious difference is the inquiry and I'm getting non-trivial shifts at the beginning of the month, that's typically age of something, and the inquiry makes sense since nothing else is moving.
Odds are it's on a 3 month or 6 month boundary to be sure, but we just don't know that conclusively yet. Still need more data, sort of like we need more data chasing that 70 or 80% installment utilization breakpoint as one non-repeatable result isn't confirmation yet.
@oilcan12 wrote:
@Thomas_Thumb wrote:
@Revelate wrote:Yuppie Foodstamp = $20 bill, so named from yuppies used to pulling out plenty of $20's from the ATM and using them copiously in restaurants late 80's early 90's.FICO doesn't appear to make a distinction between different types of installment loans other than maybe Nextgen, but it's up for debate on FICO 9 potentially.Agreed on FICO 8 = FICO 04 = an inquiry is an inquiry and remains as such for the entire year it's on there. Also agreed that my data appears to suggest that inquiries in the first 3 or possibly 6 months (where's the 5 estimate from?) but we should get close to confirming that in another 29ish days when the mortgage coded inquiry counts against me.
Oilcan12's inquieies at the time of his F09/F08 comparison were 5 months and older - His F08 scores were lower across the board than F09 scores.
That data suggests any theoretical short term impact associated with F09 had been accounted for by month 5. A shorter timeframe than 5 months (90 days or 3 months) is a more logical timeframe but his data did not include anything below 5 months.
__________________________________________________________________________________________________________________
@SouthJamaica wrote:
I had from 2 to 4 inquiries less than 3 months old on each of the 3 reports, so the distinction between more recent and less recent inquiries would not have accounted for the ~60 point difference between 8 and 9 for me.
Actually, I had a 3.5 month old Penfed HP on EQ at the time. Also, SouthJamaica says that he has 2 to 4 HPs less than 3 months old on each Bureau and his FICO 9 scores range from 50 to 73 points higher than FICO 8. I don't think that the treatment of the inqiries is the main reason for his dramatically higher FICO 9 scores, but he doesn't seem to be taking a hard hit for having multiple HPs under 3 months old.
SJ didn't specify exactly how old these inquiries were. If SJ wants to provide more precise dates on his HPs, those would be valuable data points.
TU -- 2 inquiries: 3 days; 7 days
EX -- 4 inquiries: same day, 1 day, 6 days, 35 days
EQ -- 2 inquiries: 28 days; 49 days
@Thomas_Thumb wrote:
@oilcan12 wrote:
Actually, I had a 3.5 month old Penfed HP on EQ at the time. Also, SouthJamaica says that he has 2 to 4 HPs less than 3 months old on each Bureau and his FICO 9 scores range from 50 to 73 points higher than FICO 8. I don't think that the treatment of the inqiries is the main reason for his dramatically higher FICO 9 scores, but he doesn't seem to be taking a hard hit for having multiple HPs under 3 months old.
SJ didn't specify exactly how old these inquiries were. If SJ wants to provide more precise dates on his HPs, those would be valuable data points.
Thanks for the clarification. Your data is clean and does a good job isolating a single factor, inquiries. Thus, the conclusion inquiries count less on Fico 09 relative to Fico 08 (at least over 3 months age) is sound - IMO. Your balance to loan ratio is quite low. If a high B/L ratio is penalized less on Fico 09, that further points to reduced impact of inquiries as a likely reason for your higher F09 scores.
The rather large score difference reported by SJ is not fully explainable by inquiries counting less on F09. Isolating cause(s) of his difference requires more data - right now it's speculative. Can't recall if SJ now has a high B/L ratio (car loan).
No, car loan hasn't happened yet. Utilization is perfect on single installment loan.
Revelate ha a high B/L ratio with his mortgage. A difference in weighting of aged lates and/or derogatory records is a consideration as well
I understand that SJ reported having a couple inquiries less than 90 days age.
2, 2, and 4. Details in response to Relevate.
Can't say that contradicts a 90 aging hypothesis
To my mind it does. Here I have FICO 9 scores bordering on 800, with a derogatory public record, and from 2 to 4 inquiries less than 90 days old. That certainly says to me that FICO 9 is more forgiving of recent inquiries, even inquiries within the last 90 days.
It says to me that inquiry impact and aging are less problematic to FICO 9. [Either that or FICO 9 is totally forgiving of a public record if it's going to fall off soon.... which I highly doubt is the case]
but it would be helpful to get data on potential score implications associated with monthly aging (isolating this factor is likely problematic).
Again, some type of formatted summary table would helpful for side-by-side comparison of data from various sources.
@oilcan12 wrote:I looked over my past 1B reports for a data point relevant to this discussion. I had a HP on TU on 2/25/16 for a BofA CLI. The table shows the before and after scores. They are identical.
The fact that the scores were unchanged was expected. Every credit card HP I have received TU or EQ has been buffered by 30 days or possibly ~32 days. This includes inquiries from major banks. The fact that the TU FICO 9 scores were unchanged suggests that the 30 buffer is still in effect for FICO 9.
I have absolutely 0 data points on inquiry buffering for Experian. I have too many inquiries on Experian to be able to accumulate data points.
Revelate, this is why I find it surprising that you got an immediate 80 point drop from your HP on 3/1/16. This seems to contradict all of my experiences and data. Once again, I have no data points for Experian. HPs on Experian could be treated differently.
It also seems surprising because FICO 8 seems to limit losses from inquiries to 30 points. This would be a radical departure.
Interesting, it's plausible it's something else on EX but I can't find what as explained earlier.
FICO 8 doesn't buffer CC inquiries, not sure how you've determined that unless honestly, though depending what bin you're in you may not get hit. Mortgage / Auto / Student ones are buffered and de-duped, never seen a similar buffering on my data nor any alleged FICO expert's comments.
Yeah, my data if it is inquiry related does indictate a radical departure, but it's not out of the realm of possibility it's something else but why I'm seeing major shifts in such short time periods when my files and have been steady state for 2+ years on every other model is frankly astounding. TU I can explain, obviously, but EX is behaving truly bizarrely and it's not reflected on any other score except FICO 9.
It is possible that based on the dirty bucket I am in, that + inquiries = something different than on your file. Hard to say, but it's pretty easy on my data: when my buffer period runs out on the mortgage inquiry, if I get whacked on EQ, my hypothesis has merit... if I don't, then I really don't know. I can't fathom the different buckets as TT has suggested, I've been over the files hundreds of times since I joined this forum and have been over them repeatedly as of late.
We've been having apparent issues on TU lately, maybe the other services aren't fully accurate either? Really stretching for that one given everything for me has been spot on with lender pulls personally but we have nothing else to compare FICO 9 with except here.
@oilcan12 wrote:
@Revelate wrote:
FICO 8 doesn't buffer CC inquiries, not sure how you've reached determined that unless we can't rely on the service? Mortgage / Auto / Student ones are buffered and de-duped, never seen a similar buffering on my data nor any alleged FICO expert's comments.
What is the basis for saying that FICO 8 doesn't buffer CC inquiries. My FICO 8 TU inquiries have been consistently buffered. It seems unbelievable that my TU FICO 8 CC inquiries would be buffered and yet they would not be buffered for anyone else. FICO couldn't possibly have a separate algorithm just for me.
As far as comments from Fair Isaac, if they told us everything, then no one would be trying to reverse engineer their algorithms.
I'd turn that right around, since nobody that I've read here or elsewhere over the years has ever suggested FICO 8 TU or otherwise buffered credit card inquiries.
I'll admit, I don't have enough TU data to conclusively say anything with the way it treats inquiries as I have to go way out of my way to get a TU inquiry at all whereas if I sneeze I get an EX inquiry; however, I did have a couple of non-mortgage inquiries around 8/23/15 and I ran flat TU 8 through 10/15/15 at 730... which isn't conclusive unfortunately as my score didn't materially change through there so the assumption is I didn't switch inquiry bins.
I'm not hugely fond of quoting experts or even the FICO publically released information, but it's consistent, absolutely zero mention of credit cards having a buffer in any of their published postings where they go out of their way to explain you aren't penalized for student / auto / mortgage rate shopping between the de-dupe and buffer mechanics.
One such example:
http://www.fico.com/en/blogs/risk-compliance/the-skinny-on-fico-scores-and-inquiries/
I'm not saying you're flat out utterly and absurdly wrong in your data or your analysis, just everything that we've ever found suggests your hypothesis isn't accurate.