Methods and systems for multi-credit reporting agency data modeling

• US 7,930,242 B2
• Filed: 03/12/2007
• Issued: 04/19/2011
• Est. Priority Date: 03/10/2006
• Status: Active Grant

• Current Assignee: VantageScore Solutions LLC (Equifax, Inc.)
• Original Assignee: VantageScore Solutions LLC (Equifax, Inc.)
• Inventors: Robida, Chuck, Zarikian, Lisa, Morris, Sherri
• Primary Examiner(s): Subramanian; Narayanswamy
• Assistant Examiner(s): Robinson; Kito R
• Application Number:
  US11/685,066
• Publication Number:
  US 20070255646A1
• Time in Patent Office: 1,499 Days
• Field of Search: 705/38, 705/35
• US Class Current: 705/38
• CPC Class Codes: G06Q 30/0204 Market segmentation
• CPC Class Codes: G06Q 40/00 Finance; Insurance; Tax strategies; Processing of corporate or income taxes
• CPC Class Codes: G06Q 40/02 Banking, e.g. interest calculation or account maintenance credit or loans G06Q40/03
• CPC Class Codes: G06Q 40/03 Credit; Loans; Processing thereof
• CPC Class Codes: G06Q 40/08 Insurance

Abstract

Provided are methods and systems by which a data sample can be generated in order to create a true multi-Credit Reporting Agency model, eliminating the biases and variability in traditional multi-CRA models by allowing the creation of a single multi-CRA method which requires neither alignment nor translation to be applicable on all CRAs.

Claims

1. A computer implemented method for modeling data comprising:

retrieving by the computer a source list from each of a plurality of credit reporting agencies, each said source list comprising a list of selected consumers;

merging by the computer each of the source lists into a merged source list;

retrieving by the computer observation credit data from each of the plurality of credit reporting agencies for each consumer in the merged source list, wherein the observation credit data corresponds to a first date, and wherein the observation credit data represents a credit history for each consumer;

retrieving by the computer performance credit data from each of the plurality of credit reporting agencies for each consumer in the merged source list, wherein the performance credit data corresponds to a second date, and wherein the performance credit data represents a credit history for each consumer;

adding by the computer the observation credit data and the performance credit data to the merged source list;

normalizing by the computer the observation credit data and the performance credit data across the plurality of credit reporting agencies, wherein normalizing the credit data comprises;

(i) determining at least one characteristic of the credit data, each characteristic having a corresponding characteristic definition assigned by each of at least two of the plurality of credit reporting agencies;

(ii) reviewing the at least one characteristic and its corresponding characteristic definitions, wherein the reviewing step is performed by at least two of the credit reporting agencies;

(iii) auditing the at least one characteristic and its corresponding characteristic definitions, wherein the auditing step comprises determining similarities and differences between the characteristic definitions of the at least two respective credit reporting agencies;

(iv) optionally modifying at least one characteristic definition wherein the modifying step is performed by at least one of the credit reporting agencies; and

(v) optionally repeating steps (i), (ii), (iii), and (iv) until the characteristic definition from each of the respective credit reporting agencies provides substantially consistent results; and

generating by the computer a modeling sample based on the normalized observation credit data and the normalized performance credit data.

2. The method of claim 1, wherein the first date occurs earlier in time than the second date.

3. The method of claim 2, wherein the step of adding by the computer the observation credit data and the performance credit data comprises associating the observation credit data and the performance credit data with the merged source list.

4. The method of claim 1, further comprising the step of filtering by the computer the normalized observation credit data and the performance credit data according to a desired sample.

5. The method of claim 4, wherein the desired sample is at least one of a stacked development sample, a stacked validation sample, or a parallel validation sample.

6. A system for modeling data comprising:

a memory configured for storing credit data;

a processor, coupled to the memory, wherein the processor is configured to perform the steps of;

retrieving a source list from each of a plurality of credit reporting agencies, each said source list comprising a list of selected consumers;

merging each of the source lists into a merged source list;

retrieving observation credit data from each of the plurality of credit reporting agencies for each consumer in the merged source list, wherein the observation credit data corresponds to a first date, and wherein the observation credit data represents a credit history for each consumer;

retrieving performance credit data from each of the plurality of credit reporting agencies for each consumer in the merged source list, wherein the performance credit data corresponds to a second date, and wherein the performance credit data represents a credit history for each consumer;

adding the observation credit data and the performance credit data to the merged source list;

normalizing the observation credit data and the performance credit data across the plurality of credit reporting agencies, wherein normalizing the credit data comprises;

(i) determining at least one characteristic of the credit data, each characteristic having a corresponding characteristic definition assigned by each of at least two of the plurality of credit reporting agencies;

(ii) reviewing the at least one characteristic and its corresponding characteristic definitions, wherein the reviewing step is performed by at least two of the credit reporting agencies;

(iii) auditing the at least one characteristic and its corresponding characteristic definitions, wherein the auditing step comprises determining similarities and differences between the characteristic definitions of the at least two respective credit reporting agencies;

(iv) optionally modifying at least one characteristic definition, wherein the modifying step is performed by at least one of the credit reporting agencies; and

(v) optionally repeating steps (i), (ii), (iii), and (iv) until the characteristic definition from each of the respective credit reporting agencies provides substantially consistent results; and

generating a modeling sample based on the normalized observation credit data and the normalized performance credit data.

7. The system of claim 6, wherein the first date occurs earlier in time than the second date.

8. The system of claim 7, wherein the step of adding the observation credit data and the performance credit data comprises associating the observation credit data and the performance credit data with the merged source list.

9. The system of claim 6, further comprising the step of filtering the normalized observation credit data and the performance credit data according to a desired sample.

10. The system of claim 9, wherein the desired sample is at least one of a stacked development sample, a stacked validation sample, or a parallel validation sample.

11. A non-transitory computer readable medium with computer executable instructions embodied thereon for modeling data comprising:

retrieving a source list from each of a plurality of credit reporting agencies, each said source list comprising a list of selected consumers;

merging each of the source lists into a merged source list;

retrieving observation credit data from each of the plurality of credit reporting agencies for each consumer in the merged source list, wherein the observation credit data corresponds to a first date, and wherein the observation credit data represents a credit history for each consumer;

retrieving performance credit data from each of the plurality of credit reporting agencies for each consumer in the merged source list, wherein the performance credit data corresponds to a second date, and wherein the performance credit data represents a credit history for each consumer;

adding the observation credit data and the performance credit data to the merged source list;

normalizing the observation credit data and the performance credit data across the plurality of credit reporting agencies, wherein normalizing the credit data comprises;

(i) determining at least one characteristic of the credit data, each characteristic having a corresponding characteristic definition assigned by each of at least two of the plurality of credit reporting agencies;

(ii) reviewing the at least one characteristic and its corresponding characteristic definitions, wherein the reviewing step is performed by at least two of the credit reporting agencies;

(iii) auditing the at least one characteristic and its corresponding characteristic definitions, wherein the auditing step comprises determining similarities and differences between the characteristic definitions of the at least two respective credit reporting agencies;

(iv) optionally modifying at least one characteristic definition wherein the modifying step is performed by at least one of the credit reporting agencies; and

(v) optionally repeating steps (i), (ii), (iii), and (iv) until the characteristic definition from each of the respective credit reporting agencies provides substantially consistent results; and

generating a modeling sample based on the normalized observation credit data and the normalized performance credit data.

12. The computer readable medium of claim 11, wherein the first date occurs earlier in time than the second date.

13. The computer readable medium of claim 12, wherein the step of adding the observation credit data and the performance credit data comprises associating the observation credit data and the performance credit data with the merged source list.

14. The computer readable medium of claim 11, further comprising the step of filtering the normalized observation credit data and the performance credit data according to a desired sample.

15. The computer readable medium of claim 14, wherein the desired sample is at least one of a stacked development sample, a stacked validation sample, or a parallel validation sample.