Is Equifax More Sensitive to Higher Balances?

[Anonymous]

Despite making two payments that were not credited in time, an NFCU CashRewards card I'm managing closed and reported on 8/4/18 with 70% utilization.  While EX and TU showed no change,  EQ dropped 4 points (I'm comparing 3B pulls made on 8/7/18 and 8/8/18).  I'm not concerned about the 4 points since they will be back next month but is Equifax in general more sensitive to accounts reporting high utilization that the other two CRAs?

[Anonymous]

Hard to say for sure.  The point difference is small.  I observe that the specific (individual?) utilization you quote is 70%, which is right on the cusp of where the penalty breakpoint for individual utilization is supposed to be.  Are you certain that the CC balances on all three bureaus are identical (for each of the two dates)?  Thus on the Aug 7 3B report all three bureaus say the same thing, and the same with Aug 8, with the change being between the two dates?

BTW, what tool are you using to get a 3B pull on Aug 7 and another pull on Aug 8 (with fresh data for all bureaus each time)?

[Anonymous]

---

@Anonymous wrote:

Hard to say for sure.  The point difference is small.  I observe that the specific (individual?) utilization you quote is 70%, which is right on the cusp of where the penalty breakpoint for individual utilization is supposed to be.  Are you certain that the CC balances on all three bureaus are identical (for each of the two dates)?  Thus on the Aug 7 3B report all three bureaus say the same thing, and the same with Aug 8, with the change being between the two dates?

BTW, what tool are you using to get a 3B pull on Aug 7 and another pull on Aug 8 (with fresh data for all bureaus each time)?

---

This exercise is more out of curiosity than anything else.  The account has a $1,000 CL and the balance was $689.46 or 68.9% utilization.  I've verified by checking both CCT and Experian CreditWorks that the reports contain the same information.  I tried to pay it down to below 29% before it reported but the bank mailed the payment and it took some time to arrive.  Once I realized the first payment was going to be late, I did an ACH transfer but that was also delayed because of the weekend.  So the account ended up reporting on the 4th at a much higher balance than I intended.  (No actual payment was due because these were current charges, I just didn't want it to report at 70% utilization).

In terms of credit monitoring tools, I'm not using anything outside of the norm.  I'm signed up for TrustedID (Equifax), TrueIdentity (TransUnion), and Experian Basic.  From experience I know that Experian is always the first to update my credit reports, Equifax is second, and TransUnion is always last.  TrustedID and TrueIdentity allow you to refresh your credit REPORTS daily.  So once I verify that TrustedID and TrueIdentity have the exact same information, I log in to Experian and upgrade Experian Basic to the month-long trial of Experian CreditWorks for $4.99 and get a 3B pull.  I can get another one on the 30th day when the trial expires.

But for the purpose of comparison, I signed up for the $1 CCT trial yesterday and did a 3B pull.  This pull did not include the latest update to TU.  Then today I signed up for the Experian CreditWorks trial after TrueIdentity indicated that TU had updated and did another 3B pull.  That's how I was able to see the 4-point difference after TU updated to include the $698.46 balance.  EQ already had reported the balance so the only thing that changed between the two dates was the TU update.  But still EQ lost 4 points.

[Anonymous]

I'm not sure about EQ and higher dollar/percentage balances, but I can say from my own research that EQ/TU seem to be more sensitive when it comes to number of accounts with balances compared to EX.

[Anonymous]

---

@Anonymous wrote:

I'm not sure about EQ and higher dollar/percentage balances, but I can say from my own research that EQ/TU seem to be more sensitive when it comes to number of accounts with balances compared to EX.

---

Thanks BBS.  You guys are ways more versed in the intricacies of FICO scoring so I defer to your observations. 

[Anonymous]

I probably have 1 data point that I can add later when I get home and look at my scoring spreadsheet. A few months back when at AZEO I took my AZEO card up above 50% reported utilization, then took it down to ideal again. I'll check and see how my scores reacted on each bureau and report back.

[Thomas_Thumb]

---

@Anonymous wrote:

I'm not sure about EQ and higher dollar/percentage balances, but I can say from my own research that EQ/TU seem to be more sensitive when it comes to number of accounts with balances compared to EX.

---

^ True with EQ being the most sensitive to # cards reporting.

The older EQ Fico 04 (score 5) used for mortgage loans goes bonkers when most/all cards report balances. TU Fico 04 was not as sensitive. However, since I posted a single HP on TU 9 months ago, TU Fico 04 scores have become rather skitish toward cards reporting balances.

[Anonymous]

TT, are you sure EQ is the most sensitive to number of cards reporting?  I know they are most sensitive early on (say, at 33% of cards reporting balances) but what about when you have 100% of cards reporting balances?  My initial data shows that TU is the most sensitive, then EQ as a close second followed by EX which isn't even in the same league.  I'm on my way backing down from 100% of cards with reported balances to 0% this month after the climb to 100% last month, so I'm double checking my data points along the way.

My initial find was that TU lost 15 points in going from AZEO to 100% of cards with balances, but that was a move from 850 --> 835.  There could in theory have been a few points of buffer there at 850, meaning that my loss was at least 15 points.  EQ on the other hand moved from 847 --> 836, an 11 point loss for the same move.  EX?  100% bulletproof; 847 at AZEO and 847 with 100% of cards with balances reported.

[Anonymous]

Nice conversation, guys.

Hey BBS... a good thing to bear in mind is that just because TU FICO 8 is more skittish than EQ FICO 8 (assuming that to be true) it doesn't mean that TU FICO 04 will be more skittish than EQ FICO 04 -- or TU FICO 9 than EQ FICO 9.

So as we try to distinguish about bureau difference in FICO 8 (say) we need to be careful not to believe we have shown something that applies to all models older and younger than FICO 8 (unless we have separate testing data for those).  I.e. TU is this way and EQ is that -- for all models.  It's already doubtful enough drawing conclusions given that FICO 8 has 12 different scorecards and there could variation for a how a scoring factor is handled between them.

[Thomas_Thumb]

BBS - The data I have analyzed on cards reporting balances vs Fico score was limited to:

1) Classic Fico 04/Classic Fico 98

2) Bankcard Fico 8

EQ behavior on Classic Fico 04 is quite different than EQ on Fico 8.

I don't have any Classic Fico 8 data of value for looking at score sensitivity because due to the buffer my EQ and EX Classic Fico 8 and Classic Fico 9 scores have remained constant regardless of changes in # cards reporting, individual card utilizations, aggregate utilization, or HP vs no HP.  My TU Classic Fico 8 did dip 5 points once and 2 points another time but, Classic Fico 9 held steady.

BBS has a lot more useful Classic Fico 8 data for analyzing score sensitivity.