Method and system for commercial credit analysis

• US 6,112,190 A
• Filed: 08/19/1997
• Issued: 08/29/2000
• Est. Priority Date: 08/19/1997
• Status: Expired due to Term

• Current Assignee: Citibank (South Dakota) NA (Citigroup Incorporated)
• Original Assignee: Citibank (South Dakota) NA (Citigroup Incorporated)
• Inventors: Bhat, Sandeep, Kolodziejski, Richard, Freundlich, Yehudah, Didner, Ira, Ronen, Yagil, Schoder, Joseph, Fletcher, Bruce Alan
• Primary Examiner(s): Trammell, James P.
• Assistant Examiner(s): Nguyen, Cuong H.
• Application Number:
  US08/912,866
• Time in Patent Office: 1,106 Days
• Field of Search: 705/3, 705/38, 705/35, 705/40, 705/10, 705/36, 705/44, 705/39, 705/4, 705/26, 705/37, 706/45, 706/59-60, 706/11, 706/46-47, 706/17, 706/16, 706/61, 706/10, 706/50
• US Class Current: 705/36.00R
• CPC Class Codes: G06Q 40/00 Finance; Insurance; Tax strategies; Processing of corporate or income taxes
• CPC Class Codes: G06Q 40/06 Asset management; Financial planning or analysis
• CPC Class Codes: G06Q 40/08 Insurance

Abstract

This invention describes a method and system for assimilating data, applying reasoning mechanisms, and emulating the thought processes of a credit officer for commercial credit analysis. The system aids a credit officer in the risk assessment and completion of a loan package. The system thereby improves loan turnaround time and customer service, improving loan servicing capacity, quality, and consistency of credit decisions, and reducing costs. The system includes: 1) a conventional data base management system subsystem that manages and collects data via electronic access and end-user data entry; and 2) a decision support expert system that applies a knowledge base and inference engine (collectively known as an assessment model) utilizing an evidence tree and a generalized weighting approach to analyze credit requests.

Claims

1. A method for a user to process a loan application for an applicant, comprising:

said user entering loan products requested;

said user entering data relating to said application into a data management system;

said data management system checking said data for completeness and generating preliminary results;

said data management system transmitting said data and preliminary results to an expert analysis system;

said expert analysis system assessing and recommending a loan decision using an evidence tree, wherein said data and preliminary results are propagated from a first set of nodes or factors to a second set of nodes in the evidence tree; and

generating a set of documents and transmitting applicant information from said expert analysis system to said data management system so that the data management system produces the set of documents and other forms.

2. The method of claim 1 further including:

said data management system searching and retrieving information relevant to said application from publicly available databases.

3. The method of claim 2 wherein said publicly available databases include at least one option selected from the group consisting of credit bureau agencies, UCC, litigation, and lien information.

4. The method of claim 1 further including generating a letter.

5. The method of claim 1 wherein said set of documents comprises approval documents.

6. The method of claim 1 wherein said set of documents comprises legal documents.

7. The method of claim 1 wherein said assessing and recommending includes analyzing financial records.

8. The method of claim 1 further comprising:

assigning of a debt rating;

determining a risk rating for said loan products requested;

determining an overall risk rating to the obligor; and

calculating the relationship profitability after all direct and indirect expenses.

9. The method of claim 1 wherein said assessing and recommending includes identifying sources of repayment.

10. The method of claim 1 wherein said data relating to an application comprises demographic, business, banking, products, industry, business financial, personal, and personal financial information, and subjective evaluations.

11. The method of claim 1 wherein the assessing and recommending comprises:

preprocessing prior to execution to align evidence tree elements based on the profile of said loan application;

analyzing inputted data in preparation for input to the evidence tree elements;

inputting said analyzed inputted data into the evidence tree elements;

evaluating the evidence tree elements based on the weights assigned, the rules, and the inputted data;

producing an explanation of strengths and weaknesses of the loan application based on inputs to and results of evidence tree analysis; and

making a loan recommendation.

12. The method of claim 11 wherein said loan recommendation includes at least one option selected from the group consisting of approve, decline, defer, refer, and neutral, and includes a grading of at least one option selected from the group consisting of excellent, good, fair, marginal, poor, and very poor, and includes an overall rating.

13. The method of claim 11 wherein said explanations comprises narrative text.

14. The method of claim 11 wherein said strengths are listed prior to said weaknesses for an approved loan application.

15. The method of claim 11 wherein said weaknesses are listed prior to said strengths for a declined loan application.

16. A method for an expert analysis system to assist a user with making a loan decision based on inputted data, comprising:

preprocessing to align evidence tree elements and assign weights based on the profile of said loan application;

analyzing the inputted data in preparation for input to the evidence tree elements;

inputting said analyzed inputted data into the evidence tree elements;

assigning values to some of the inputted data based on a scale;

evaluating the evidence tree elements based on the weights assigned and the values by multiplying some of the values by some of the weights to create weighted values that are propagating from a first node to a second node;

producing an explanation of strengths and weaknesses of the loan application based on inputs to and results of the evidence tree analysis; and

making a loan recommendation.

17. The method of claim 16 wherein said evaluating includes adding additional values assigned to evidence tree factors and nodes.

18. The method of claim 16 wherein said evaluating includes selecting a largest numeric value assigned to one of the evidence tree factors and nodes.

19. The method of claim 18 further including applying a numeric modifier.

20. The method of claim 19 wherein said modifier is a percentage.

21. The method of claim 16 wherein said evaluating includes selecting a smallest numeric value assigned to one of the factors and nodes.

22. The method of claim 16 further comprising saving data from said loan application.

23. The method of claim 16 further comprising:

inputting data for multiple scenarios;

producing results for multiple scenarios;

presenting to the user said results; and

producing approval documents for said multiple scenarios.

24. The method of claim 16 further comprising linking to a database system for data extraction.

25. The method of claim 16 wherein said loan recommendation includes at least one option selected from the group consisting of Approve, Decline, Neutral, Refer, and Defer.

26. The method of claim 16 further includes:

tagging selected factors of the evidence tree;

collecting said tagged factors;

combining said tagged factors; and

assigning narrative information to said combined tagged factors.

27. The method of claim 16 further including:

identifying a decision tree element that is not applicable to the evidence tree analysis evaluation; and

replacing said inputted data in evidence tree elements with a textual symbol.

28. The method of claim 27 wherein said textual symbol is "N/A".

29. The method of claim 16 further comprising:

identifying risk acceptance criteria and underwriting guidelines for evidence tree elements; and

providing a flag within the system identifying said risk acceptance criteria and said underwriting guidelines as at least one option selected from the group consisting of pass and fail.

30. The method of claim 29 wherein the flag is included in a factor.

31. The method of claim 29 wherein said flag comprises a highlighted symbol.

32. The method of claim 16 further comprising:

identifying an exceptionally negative factor within the evidence tree; and

providing a list of extreme negative factors that includes said identified extreme negative factor.

33. The method of claim 16 further comprising:

identifying an exceptionally positive factor; and

providing a list of exceptionally positive factors that includes said identified exceptionally positive factor.

34. The method of claim 16 further comprising:

identifying that an overriding rule has been triggered; and

providing a flag to the user that an overriding rule has been triggered.

35. The method of claim 16 further comprising:

determining by said expert system that no data has been entered for one of the evidence tree elements; and

providing to the user an indication that no data was entered for the one evidence tree element for which no data has been entered.

36. The method of claim 16 further comprising tagging factors.

37. The method of claim 36 wherein said tagging includes color tagging.

38. The method of claim 36 wherein said tagging includes verbal tagging.

39. The method of claim 36 wherein said tagging includes textual tagging.

40. The method of claim 36 wherein said tagging includes symbol tagging.

41. The method of claim 36 wherein said tagging includes numeric tagging.

42. The method of claim 16 further comprises

allowing the user to view color tags, verbal tags, textual, symbol, and numeric tags, and real-time propagation of the evaluating step.

43. The method of claim 16 further comprising:

tagging each evidence tree element for which an explanations is available; and

allowing the user to select an option to view said explanation.

44. The method of claim 16 further comprises

providing a list of factors that were not evaluated where an incomplete set of input data has been inputted by said user.

45. The method of claim 16 further comprises:

identifying qualitative factors in the evidence tree;

combining said qualitative factors; and

providing said user with the total of said qualitative factors and a range of possible qualitative factors.

46. The method of claim 45 wherein said qualitative factors comprise subjective factors.

47. The method of claim 45 wherein said qualitative factors include identified factors in an evidence tree.

48. The method of claim 45 wherein said combining includes summing numeric values associated with evidence tree branches.

49. A system for a user to process a loan application by an applicant, comprising:

means for entering data relating to said application into a data management system;

means for said data management system to check said data;

means for said data management system to generate relevant documents for said applicant relating to said application;

means for said data management system to transmit said data to an evidence tree;

means for said evidence tree to assess and recommend a loan decision by propagating data from a first node or factor to a second node in the evidence tree; and

means for said evidence tree to transmit applicant information to said data management system so that the data management system may generate a set of relevant documents.

50. The system of claim 49 further comprising:

means for searching and retrieving within said data management system of publicly available databases for information relevant to said loan application, including UCC, litigation, and lien information relevant to said application.

51. The system of claim 49 further comprising generating a letter.

52. The system of claim 49 wherein said means for checking, generating, searching, assessing, filing, and recommending comprise computing means.

53. A method for processing an application comprising:

entering applicant data into a processing system;

organizing the applicant data into formatted data by the processing system

generating preliminary results by the processing system;

forwarding the preliminary results to an expert analysis system comprised of an evidence tree;

accessing at least one database to obtain additional data;

generating second results by the expert system based upon the formatted data, preliminary results and the additional data using the evidence tree containing a plurality of nodes and factors and at least one of the following techniques;

propagating a value from a first node or a first factor to a second node;

propagating a weighted value from a third node or a second factor to a fourth node;

forwarding the results to the processing system; and

producing documents necessary to manage a financial transaction by the processing system wherein some of the results from the expert analysis system are contained in the documents.

54. The method of claim 53 further comprising:

identifying suggested alternative possible evaluation outcomes for nodes and factors; and

displaying said alternative outcomes.

55. The method of claim 53 further comprising displaying said nodes and factors without alternative outcomes.

56. The method of claim 53 further comprising the step of displaying said nodes and factors without tags and flags.

57. The method of claim 53 wherein said factors comprise qualitative evaluation by said user, including at least one option selected from the group consisting of satisfactory and unsatisfactory for supplier risk, and at least one option selected from the group consisting of high, medium, and low for customer stability.