SCORECARDS?

[Thomas_Thumb]

Thanks for the pdf format - it has been added to my electronic library.

Really, the "white paper" is more of a marketing tool to sell Fico scorecard services to businesses - IMO. The fuzzy screenshots I pasted from the slide share are meant to be informative in general terms to illustrate segmentation to create multiple scorecards and how  bins within scorecards are used.

The segmentation approach is a key illustration in understanding how Fico scorecards are developed.

I used a segmentation technique to come up with the known Fico 8, 12 scorecard count, using the factors listed in Liz's book.

[Anonymous]

Good stuff here, folks.  I truly appreciate it.  At the moment it may not be in my understanding, but it's a treat to have available for the moment I am ready, which often comes on a whim.  

[Thomas_Thumb]

Here is another interesting article on segmentation from Fico along with an image capture from the white paper.

http://www.fico.com/en/node/8140?file=9737

[RobertEG]

Segmentation is clearly a standard tool in scorecard development, and is used in the FICO algorithms.

However, since Fair Isaac made their decisiion years ago to forgo patent protection and attempt to extend their propriietary protection by the trade secret route, they are not going to disclose adequate information to permit reverse engineering of their trade secret algoriths.

Thus, details to the extent necessary to fully understand their operation will intentionally remain in Brother Isaac's vault.

The extensive literature on scorecard segmentation of various scoring categories generally follows the key "thngs" identified in DIYCredit's orignal post.

[Thomas_Thumb]

In general, reverse engineering is done by observation of cause/effect and measurement of known "critical criteria".

If info is freely given, there would be no need to reverse engineer. At the end of the day, existing Fico models are mature - have plateaued in their life cycle relative to market penetration.

Competition in the marketplace necessitates change to "capture" additional population segments and to maximize profitability by offering additional value add to business partners - while also courting consumer revenue through services such as credit monitoring and alerts.

In general, it is rather difficult to protect data analysis techniques as "patentable inventions".

[Anonymous]

Thom, I just wanted to say this is a very enlighting concept for me, as I didn't realize all of the different scorecards.  I want to provide you with my information as a point of reference.  I believe according to your chart I would be in either scorecard 11 or 12, I am not exactly sure if you only need to meet 1 category as H.  Scores come from a myfico 3b report from May 2016.  If you want any of the other Fico score versions I can provide them. 

EQ fico 8 Score: 704

Derog: Forclosure (DOFD 10/2010)

Oldest Acct: Opened 9/2004

Youngest Acct: Opened 3/2016

Total Accounts: 6

TU fico 8 Score: 669

Derog: Forclosure (DOFD 10/2010)

Oldest Acct: Opened 6/2005

Youngest Acct: Opened 3/2016

Total Accounts: 5

TU fico 8 Score: 663

Derog: Forclosure (DOFD 10/2010)

Oldest Acct: Opened 6/2005

Youngest Acct: Opened 3/2016

Total Accounts: 5

[RonM21]

This is great info...thanks

[Thomas_Thumb]

---

@Anonymous wrote:

Thom, I just wanted to say this is a very enlighting concept for me, as I didn't realize all of the different scorecards.  I want to provide you with my information as a point of reference.  I believe according to your chart I would be in either scorecard 11 or 12, I am not exactly sure if you only need to meet 1 category as H.  Scores come from a myfico 3b report from May 2016.  If you want any of the other Fico score versions I can provide them. 

EQ fico 8 Score: 704

Derog: Forclosure (DOFD 10/2010)

Oldest Acct: Opened 9/2004

Youngest Acct: Opened 3/2016

Total Accounts: 6

TU fico 8 Score: 669

Derog: Forclosure (DOFD 10/2010)

Oldest Acct: Opened 6/2005

Youngest Acct: Opened 3/2016

Total Accounts: 5

TU EX fico 8 Score: 663

Derog: Forclosure (DOFD 10/2010)

Oldest Acct: Opened 6/2005

Youngest Acct: Opened 3/2016

Total Accounts: 5

---

Thanks for the data. Given the age of your foreclosure, I would assign your EQ scorecard an "H" rating relative to the table. Age is over 3 years, you only mention one derog and you list more than 5 accounts. Really not sure why your EQ is 40 points higher than TU/EX, assuming the 3rd group above represents EX. The only difference is # accounts (5 compared to 6). What is the 6th account and do you have differences in # inquiries under 12 months age among the CRAs?

Looks like an adjustment to scorecard 11 ceiling might be in order.

[Revelate]

---

@Thomas_Thumb wrote:

---

@Anonymous wrote:

Thom, I just wanted to say this is a very enlighting concept for me, as I didn't realize all of the different scorecards.  I want to provide you with my information as a point of reference.  I believe according to your chart I would be in either scorecard 11 or 12, I am not exactly sure if you only need to meet 1 category as H.  Scores come from a myfico 3b report from May 2016.  If you want any of the other Fico score versions I can provide them. 

EQ fico 8 Score: 704

Derog: Forclosure (DOFD 10/2010)

Oldest Acct: Opened 9/2004

Youngest Acct: Opened 3/2016

Total Accounts: 6

TU fico 8 Score: 669

Derog: Forclosure (DOFD 10/2010)

Oldest Acct: Opened 6/2005

Youngest Acct: Opened 3/2016

Total Accounts: 5

TU EX fico 8 Score: 663

Derog: Forclosure (DOFD 10/2010)

Oldest Acct: Opened 6/2005

Youngest Acct: Opened 3/2016

Total Accounts: 5

---

Thanks for the data. Given the age of your foreclosure, I would assign your EQ scorecard an "H" rating relative to the table. Age is over 3 years, you only mention one derog and you list more than 5 accounts. Really not sure why your EQ is 40 points higher than TU/EX, assuming the 3rd group above represents EX. The only difference is # accounts (5 compared to 6). What is the 6th account and do you have differences in # inquiries under 12 months age among the CRAs?

Looks like an adjustment to scorecard 11 ceiling might be in order.

---

Well from the posted data, the account which isn't on EX/TU is the oldest account, if the AAOA of the others is comparitively low (youngest account 2 months ago so presumably not at credit file complete) might be a non-trivial change in that though 40 points is a big disparity; however, 10 points for AAOA and 30 points FICO feng shui handwaving resolution would be reasonable.

[RobertEG]

Thomas, can you explain a bit on what the scorecard designations, such as H-H-L, mean when applied to actual files?

What does High and Low refer to?