Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

A method to determine a relative ranking of customers credit worthiness

IP.com Disclosure Number: IPCOM000015032D
Original Publication Date: 2001-Nov-26
Included in the Prior Art Database: 2003-Jun-20
Document File: 4 page(s) / 84K

Publishing Venue

IBM

Abstract

Credit Scoring is one of the worlds most commonly used method to decide whether a financial institution should grant a credit product such as credit card, loan services or mail order purchase, to their applicants. This result of this assessment method is a score that enables financial institutions to determine the applicants' credit worthiness. In general, the higher that credit score, the more trust worthy that this applicant is. However, before this score can be calculated, a credit scorecard must be built place prior to the scoring exercise. A credit scorecard is a decision making tool which is developed by means of statistical analyses. The analytical procedures involved in developing such credit scorecards are: 1) Data Collection and Sampling 2) Scorecard Definition and Scorecard Splitting 3) Good/Bad Definition

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 29% of the total text.

Page 1 of 4

A method to determine a relative ranking of customers credit worthiness

   Credit Scoring is one of the worlds most commonly used method to decide whether a financial institution should grant a credit product such as credit card, loan services or mail order purchase, to their applicants. This result of this assessment method is a score that enables financial institutions to determine the applicants' credit worthiness. In general, the higher that credit score, the more trust worthy that this applicant is. However, before this score can be calculated, a credit scorecard must be built place prior to the scoring exercise. A credit scorecard is a decision making tool which is developed by means of statistical analyses. The analytical procedures involved in developing such credit scorecards are:

1) Data Collection and Sampling
2) Scorecard Definition and Scorecard Splitting
3) Good/Bad Definition
4) Characteristics Definition
5) Attributes Definition
6) Fine and Coarse Classing
7) Characteristics Selection
8) Scorecard Building

Step 1 and 2 are standard procedural steps in the Credit Scorecard building process and do not involve any analytical effort. Step 3 onward is a series of analytical work that are sequentially linked together in the order specified above and no single step can be individually executed or missed. It is vitally important to start the Good/Bad Definition as soon as the data is available after Step 2. A good scorecard therefore stems from a scrupulous and unbiased measurement and derivation of the Good/Bad Definition. In building any type Credit Scorecard, whether it be an Application, Behavioural or Collection Scorecard, this Good/Bad Definition step is extremely important. Its primary objective is to distinguish the "GOOD" customers from the "BAD" customers so that the scorecard is taught the pattern of the Good and Bad customers and to make a sound and unbiased judgement in future when applied to live sample. Good customers in general and by business rules definition, are those who do not own financial institutions such as banks and money lenders substantial amount of money and that these institutions do not need to incur any extra cost in recovering the debt. In practice, take credit card or loan product for example, customers who fall in arrears for 3 months or more are considered as Bad customers. There are of course many other definitional boundaries to confine those who do not have 3 months or more in arrears but are constant late payers or debt dodger. The bottom line is what are those definitional boundaries, how are they derived, why the 3 month threshold is used as the Good/Bad separation line so subjectively. Could it be 2 months or 6 months? This invention attempts to build an objective comparison method to determine who should be Good and who should be Bad based on all qualified customers' historical performance. This invention will replace the Good/Bad Definition step mentioned above.

The traditional approach o...