cancel
Showing results for 
Search instead for 
Did you mean: 

Head Scratcher. The Mystery of FICO Formula.

tag
ztnjpv
Established Contributor

Head Scratcher. The Mystery of FICO Formula.

There is more to the formula than meets the eye. Consider the recent journey of my Equifax report and score:

 

4/10/12: FICO Score 718

 

Payment History: Good, Amount of debt: Very Good, Length of Credit History: Very Good, Amount of New Credit: Not Good.

 

I have a public record AND a serious delinquency (paid CC charge off from 2006/2007)

AAofA: 7 years

Time since last delinquency 5 yrs, 5 ms

Number of accounts with red flag: 2 (doesn't include the public record)

Inquiries: 3

 

Accounts [?]Accounts with balances [?]Accounts opened in past year [?]Recent inquiries [?]Collections [?]Public Records [?]

15 
2 
3 
4 
0 
1

 


 

4/25/12: FICO Score 719

 

Payment History: Very Good, Amount of debt: Very Good, Length of Credit History: Very Good, Amount of New Credit: Bad.

 

I have a public record from 2006

AAofA: 7 years

Time since last delinquency 5 yrs, 7 ms

Number of accounts with red flag: 1(doesn't include the public record)

Inquiries: 4

 

Accounts [?]Accounts with balances [?]Accounts opened in past year [?]Recent inquiries [?]Collections [?]Public Records [?]

15 
3 
4 
4 
0 
1

 

 

So what changed in between those dates? 

 

Well, the CC charge off from 2006 was deleted and an authorized user account was added (with a balance) on a card opened 2 months ago by my wife.

 

My verizon baddie was updated with no difference....but it was updated. 4/2012 is now the most recent update. Also, this account doesn't show up as a red flag on either report. However, in between the dates, Verizon accidentally changed this verizon account from "Closed at Consumer's Request" to "Collection" and my score dropped to 691 even though that CC charge off had come off at the same time. But then on 4/25/12, it went back to "Closed at Consumer's Request" but with an April 2012 updated status. 

 

Find this puzzling. When that CC Charge off came off my TU FICO, my score jumped from 723 to 781. Other differences in that report are 1. more closed postive accounts, 2. no public record, 3. no April update from Verizon. But it still has the one red flag. That red flag on both EQ's and TU's report is a 30 day late payment from 2006 to VW Credit. 

 

Also: On the day I had a 718 FICO, my EQ Score was 701. But now both my FICO and EQ Score are 719! I know the formulas differ a little but it's the first time my EQ Score was NOT at least 15 points lower than my real FICO. BTW, when my FICO temporarily dropped to 691, my EQ Score dropped to 676. So when my EQ Score shot back up to 719 after the correction by Verizon, I was expecting to see a FICO of about 735 give or take. 

 

Strange.


Start (Sept 2011): low-mid 600s. NOW: TU FICO: 801, EQ FICO 808, EX FICO 798 (PSECU). Goal: Achieved! Now Maintain!
Message 1 of 4
3 REPLIES 3
llecs
Moderator Emeritus

Re: Head Scratcher. The Mystery of FICO Formula.

The FICO formula between TU and EQ isn't the same. IME, TU weighed baddies more than EQ and EQ weighed util more than TU. Anyway, your TU is lacking the PR and you are in a different scoring bucket on TU because it isn't reporting the PR. Since the baddie was really the only thing weighing you down (baddie-wise), you stood to see a larger gain.

 

Don't compare EQ's FAKO ("Equifax Credit Score") with FICO. I've personally have seen 100 point differences when pulled the exact same day.

Message 2 of 4
ztnjpv
Established Contributor

Re: Head Scratcher. The Mystery of FICO Formula.

Thanks llecs,

 

That point about a different FICO formula for each report is news to me. I always assumed FICO was FICO and that the difference in scores was always because of the contents of the reports themselves. 

 

Your next point about  "scoring buckets"  reminds me of an article by attorney, in which he described how a client's score strangley dropped because he moved in a different comparable peer group...or something to that effect....when certain bad things were removed. It may seem confusing as I describe it to some readers but he made make sense. It has to do with the thresholds and levels that consumers fall into based on their scores. I thought about this when my EQ score did not increase as much as my TU when the CO come off. 

 

On the last part, I understand what you are saying but I am just referring to my experience with the information in my particular report. That point gap I mentioned had been pretty consistent up until this matter at hand. And again, I suspected that there were larger changes at play in my overall place within the consumer hierarchy...or buckets. I feel this aspect of the scoring model doesn't get enough discussion in the form of articles by credit experts. 

Start (Sept 2011): low-mid 600s. NOW: TU FICO: 801, EQ FICO 808, EX FICO 798 (PSECU). Goal: Achieved! Now Maintain!
Message 3 of 4
RobertEG
Legendary Contributor

Re: Head Scratcher. The Mystery of FICO Formula.

Yeah, making the assumption that FICO is a uniform scoring algorithm can lead to total confusion.

Fair Isaac markets dozens of scoring algorithms, each serving the requested needs of their customers.  Each CRA licenses it own flavor.

Different industry segments, such as auto and mortgage creditors, have their own risk analysis factors, differently weighted to serve their needs.

 

Within the algorithm itself, different weightings are made for different factors based on some mysterious overall categorization of a consumer's credit file..... commonly referred to as bucketing.

The effect of derogs, for example, can vary based on whether the overall file has other major derogs remainging (a "dirty" file), or is otherwise "clean" of major derogs.

The 35%/30%/15%/10%/10% major category weigtings are not fixed in stone, but apparently vary based on bucketing.  It goes on and on for different scoring categories.

 

Fair Isaac has never disclosed the proprietary trade secret innards of scoring, so we can only guess as to how the algorithms are affected based on such factors.

All we can say for sure is that it is not a one shoe fits all feet approach.

Message 4 of 4
Advertiser Disclosure: The offers that appear on this site are from third party advertisers from whom FICO receives compensation.