cancel
Showing results for 
Search instead for 
Did you mean: 

Credit Scoring and "model building"

tag
veracious
Established Contributor

Credit Scoring and "model building"

I read an  older article about this topic. The article mentioned that there are

  "30 or so possible characteristics"   that are derived from our reports.

I've read here at My FICO that there are 12 (so-called)  scoring buckets.

So if gender, race, religion, occupation??, etc. are not considered,

what could these  "30 or so possible characteristics"  be?

The person being interviewed in the article was a chief analyst manager from  Fair Isaac Co.

 BTW: Here's another one_

I saw this on www. fico.com   May 21,2010  Fico patents 8 new analytic Techniques for Scoring....

 

Will any of these things bring about changes; or are they already in effect ?

                                                                                                                                                      

edited to add:     This article also stated that a new model is built every  two years  based on

the previous 2year profilesof millions of consumers.

I was wondering if  the users of  My FICO are an experiment in progress?

 

Thanks for any replies

 

_________________________________________________
"You may never know what results come of your actions,
but if you do nothing, there will be no result" ~ Mahatma Gandhi
Message 1 of 6
5 REPLIES 5
Anonymous
Not applicable

Re: Credit Scoring and "model building"


@veracious wrote:

I read an  older article about this topic. The article mentioned that there are

  "30 or so possible characteristics"   that are derived from our reports. 


I'm not sure which article you're referring to, but my guess is that they're referring to the FICO Reason Codes.  These are public information, but I think they are one of the most useful tools for learning how to manage your FICO score.  You can google them, or find them sticked at the top of this board as 

I think knowing the reason codes is the bee's knees.  Good question, veracious!
Message 2 of 6
veracious
Established Contributor

Re: Credit Scoring and "model building"

Thanks, beamMEup.           Good answer.

 

My other  question would be: Since I also read that there are two types (maybe more ) of

credit scoring models:   1.Judgemental  Scoring  Model

                                            2.Statistical Scoring Model

Can I guess that one model would be a model used  strictly for on-line  CC apps.,

and the other would be used for face-face apps?

Since banks and CU's rely on more than just our raw credit scores,  I would appreciate

any light you can shed on the subject.

 

Thanks.             veracious

_________________________________________________
"You may never know what results come of your actions,
but if you do nothing, there will be no result" ~ Mahatma Gandhi
Message 3 of 6
RobertEG
Legendary Contributor

Re: Credit Scoring and "model building"

Most who generate credit scoring keep their algorithms proprietary, to there is little in the published literature describing how “buckets” specifically affect scoring.

However, First USA Bank applied for and received a patent in 2001 describing in detail the categories/algrithms that they proposed. at ;least bacl then, to score consumer credit risk..  It is U.S. Patent No. 6,202,053

It describes their scoring algorithm based on “buckets.”  An interesting read.

In their published methodology, a first cut is made based on whether the credit file is “thick” or “thin.”  In general, a thin file is less than three trade lines, or no credit card.

A second major branch is based on delinquencies.  Files are classified as “clean” or “dirty”, based on the absence or presence of a major delinquency in the file, which is defined as a late of 60+, or a CO, CA, or PR.

A third major branch is %util on revolving accounts, with various sub-branches thereunder, depending upon the %util level.

A fourth branch is length of credit history.  They use less than 10 years, or 10+ years as a primary breakdown.

There are sub-branches off of these major branches, which in all, match approximately the 12 major buckets that are reportedly used by FICO.

 

This description is, of course, not published by Fair Isaac, and is not the precise FICO methodology, but it is one of the few published, detailed descriptions of how a major credit scoring corporation views the breakdown of multiple scoring algorithms based on scoring "buckets."  I dont think you will find a more complete, published discussion by any creditor of how they view scoring buckets than is disclosed in this patent.I It is exremely rare for one to attempt to patent a scoring algorithm, for thar removes the information from the trade secret realm, nd puts it  into the public disclosure realm.

You can get a copy on the Patent Office website, at  uspto.gov.

Message 4 of 6
RobertEG
Legendary Contributor

Re: Credit Scoring and "model building"

Whether a potential creditor relies primarily, or in some cases, only on FICO scoring in making credit decisions, is based on a weighing of credit investigation costs and delays in doing a more detailed review, against the quick and simple purchase by them of your CR and score.  If a creditor denies credit based primarily on a FICO score, they are required, under the FCRA, to notify you of the factors that impeded your approva.

For example, if you apply for a low CL card, the credtior gets thousands of such applications daily.  They are usuaklly not going to do a full, manual review of each applicant, and rarely ask for things not included in your CR, such as income, total debts, total assets, etc., if  you meet the FICO threshold they arbitrarily set as their comfort zone.

Higher end cards might also additionally ask simply for income. 

When you come to the higher principal credit extensions, such as auto or mortgage loans, you can expect a more intensive review.  They will spend the time and money to protect their higher risk.

Message 5 of 6
veracious
Established Contributor

Re: Credit Scoring and "model building"

Thanks RobertEG,

 

Once again you have provided insightful and relevant answers to my question.

I appreciate your time and your sharing of knowledge.

 

 

         veracious...

 

_________________________________________________
"You may never know what results come of your actions,
but if you do nothing, there will be no result" ~ Mahatma Gandhi
Message 6 of 6
Advertiser Disclosure: The offers that appear on this site are from third party advertisers from whom FICO receives compensation.