Method and system for enhancing credit line management, price management and other discretionary levels setting for financial accounts

• US 20040111363A1
• Filed: 05/21/2003
• Published: 06/10/2004
• Est. Priority Date: 11/18/2002
• Status: Active Grant

• Current Assignee: JP Morgan Chase Bank, N.A. (JP Morgan Chase & Co)
• Original Assignee: First USA Bank N.A. (JP Morgan Chase & Co)
• Inventors: Lau, Tak Wing, Wang, Hui, Nair, Suresh K., Ma, Lizhi, Pederson, Shane P., Trench, Margaret S.
• Granted Patent: US 7,720,761 B2
• Time in Patent Office: Days
• US Class Current: 705/39
• CPC Class Codes: G06Q 20/10 specially adapted for electronic funds transfer [EFT] systems; specially adapted for home banking systems
• CPC Class Codes: G06Q 20/108 Remote banking, e.g. home banking
• CPC Class Codes: G06Q 20/403 Solvency checks
• 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

Abstract

A method and system is disclosed for enabling the accurate determination of price points (APRs), credit limits, and other discretionary levels for each cardholder that maximize Net Present Value (NPV) for the portfolio, given constraints on quantities such as risk of default. In accordance with one embodiment, the present invention uses a Markov Decision Process (MDP) methodology to generate a simplified transition matrix representative of the potential state transitions for account holders. This model applies account level historical information on purchases, payments, profitability and delinquency risk to make these decisions. In addition, a data structure is disclosed constructed to implement a transition matrix computationally in different sizes.

Claims

1. A method for managing financial accounts, comprising the steps of:

collecting account data for a plurality of financial accounts, where the account data are associated with a plurality of variables;

establishing at least one control variable and at least one behavior variable based on the plurality of variables;

determining a plurality of interval breakpoints for the at least one control variable and the at least one behavior variable;

defining a plurality of account states based on the at least one control variable, the at least one behavior variable and the plurality of interval breakpoints;

forming a transition matrix based at least in part on the account data and the plurality of account states;

generating a Markov Decision Process (MDP) model based at least in part on the transition matrix; and

using the Markov Decision Process model to effect a change in the at least one control variable resulting in a desired effect on the plurality of account states.

2. The method according to claim 1, where the plurality of financial accounts are bankcard accounts and the at least one control variable comprises at least one of:

(i) credit lines;

or

(ii) annual percentage rates (APRs).

3. The method according to claim 1, where the account data comprise time-series data on at least one of:

(i) credit bureau variables;

(ii) credit bureau scores; and

(iii) bankcard account performance.

4. The method according to claim 2, where the change in the at least one control variable for the desired effect comprises at least one of the following actions:

(i) change the credit line;

(ii) change the annual percentage rate (APR);

(iii) change both the credit line and the annual percentage rate (APR);

or

(iv) making no changes.

5. The method according to claim 2, where the desired effect is associated with at least one of the bankcard accounts'"'"':

(i) net cash flow (NCF);

(ii) balance migration;

or

(iii) risk of credit losses.

6. The method according to claim 1, further comprising the steps of:

performing a regression tree analysis on the plurality of variables;

establishing the at least one behavior variable based on the regression tree analysis; and

determining the plurality of interval breakpoints for each of the at least one behavior variable based on the regression tree analysis.

7. The method according to claim 1, further comprising the steps of

applying business rules to an output of the Markov Decision Process model;

and

generating a policy lookup table based on the business rules and the output of the Markov Decision Process model.

8. The method according to claim 1, further comprising simplifying the transition matrix based on one or more of the following techniques:

(i) creating a rectangular transition matrix with rows for the at least one control variable and the at least one behavior variable, and columns for the at least one behavior variable only;

(ii) combining at least one row of a transition matrix with an adjoining row, where the at least one row is characterized with low-frequency transitions;

(iii) combining at least one column of a transition matrix with an adjoining column, where the at least one column is characterized with low-frequency transitions; and

(iv) re-indexing a transition matrix to correctly capture all past and future account transitions.

9. The method according to claim 1, further comprising validating and supplementing the output of the Markov Decision Process model through a Monte Carlo simulation, where the Monte Carlo simulation is based at least partially on the account data.

10. A data structure for implementing a transition matrix computationally in a plurality of sizes, the data structure comprising:

a plurality of control branches, where each of the plurality of control branches is a member of a first bi-directionally linked list comprising;

a plurality of control variables, a bi-directionally linked list of behavioral state nodes, and an array of action-specific destinations of the plurality of control variables;

a plurality of behavior nodes, where each of the plurality of behavior nodes is a member of a second bi-directionally linked list comprising;

a plurality of behavioral state variables, a plurality of arrays of financial metrics and simulation statistics, an array of action-specific destinations of the plurality of behavioral state variables, and a bi-directionally linked list of all behavioral transitions; and

a plurality of transition probability nodes, where each of the plurality of transition probability nodes is a member of a third bi-directionally linked list comprising;

a plurality of destination behavioral state variables, a list of destination pointers for behavior nodes that have non-zero transition probabilities, and a list of transition probability values of all destination points.

11. A computer readable medium having code for causing a processor to manage financial accounts, the computer readable medium comprising:

code adapted to collect account data for a plurality of financial accounts, where the account data are associated with a plurality of variables;

code adapted to establish at least one control variable and at least one behavior variable based on the plurality of variables;

code adapted to determine a plurality of interval breakpoints for the at least one control variable and the at least one behavior variable;

code adapted to define a plurality of account states based at least in part on the at least one control variable, the at least one behavior variable and the plurality of interval breakpoints;

code adapted to form a transition matrix based at least in part on the account data and the plurality of account states;

code adapted to generate a Markov Decision Process (MDP) model based at least in part on the transition matrix; and

code adapted to use the Markov Decision Process model to effect a change in the at least one control variable resulting in a desired effect on the plurality of account states.

12. The computer readable medium according to claim 11, where the plurality of financial accounts are bankcard accounts and the at least one control variable comprises at least one of:

(i) credit lines;

or

(ii) annual percentage rates (APRs).

13. The computer readable medium according to claim 11, where the account data comprise time-series data on at least one of:

(i) credit bureau variables;

(ii) credit bureau scores; and

(iii) bankcard account performance.

14. The computer readable medium according to claim 12, where the change in the at least one control variable for the desired effect comprises at least one of the following actions:

(i) change the credit line;

(ii) change the annual percentage rate (APR);

(iii) change both the credit line and the annual percentage rate (APR);

or

(iv) making no changes.

15. The computer readable medium according to claim 12, where the desired effect is associated with at least one of the bankcard accounts'"'"':

(i) net cash flow (NCF);

(ii) balance migration;

or

(iii) risk of credit losses.

16. The computer readable medium according to claim 11, further comprising:

code adapted to perform a regression tree analysis on the plurality of variables;

code adapted to establish the at least one behavior variable based on the regression tree analysis; and

code adapted to determine the plurality of interval breakpoints for each of the at least one behavior variable based on the regression tree analysis.

17. The computer readable medium according to claim 11, further comprising

code adapted to apply business rules to an output of the Markov Decision Process model;

and

code adapted to generate a policy lookup table based on the business rules and the output of the Markov Decision Process model.

18. The computer readable medium according to claim 11, further comprising code adapted to simplify the transition matrix based on one or more of the following techniques:

(i) creating a rectangular transition matrix with rows for the at least one control variable and the at least one behavior variable, and columns for the at least one behavior variable only;

(ii) combining at least one row of a transition matrix with an adjoining row, where the at least one row is characterized with low-frequency transitions;

(iii) combining at least one column of a transition matrix with an adjoining column, where the at least one column is characterized with low-frequency transitions; and

(iv) re-indexing a transition matrix to correctly capture all past and future account transitions.

19. The computer readable medium according to claim 11, further comprising code adapted to validate and supplement the output of the Markov Decision Process model through a Monte Carlo simulation, where the Monte Carlo simulation is based at least partially on the account data.

20. A system for managing financial accounts, the system comprising

a data collection module for collecting account data for a plurality of financial accounts, where the account data are associated with a plurality of variables;

an establishment module for establishing at least one control variable and at least one behavior variable based on the plurality of variables;

a determination module for determining a plurality of interval breakpoints for the at least one control variable and the at least one behavior variable;

a definition module for defining a plurality of account states based at least in part on the at least one control variable, the at least one behavior variable and the plurality of interval breakpoints;

a formation module for forming a transition matrix based at least in part on the account data and the plurality of account states;

a generation module for generating a Markov Decision Process (MDP) model based at least in part on the transition matrix; and

a decision module for using the Markov Decision Process model to effect a change in the at least one control variable resulting in a desired effect on each of the plurality of account states.

21. The system according to claim 20, where the plurality of financial accounts are bankcard accounts and the at least one control variable comprises at least one of:

(i) credit lines;

or

(ii) annual percentage rates (APRs).

22. The system according to claim 20, where the account data comprise time-series data on at least one of:

(i) credit bureau variables;

(ii) credit bureau scores; and

(iii) bankcard account performance.

23. The system according to claim 21, where the change in the at least one control variable for the desired effect comprises at least one of the following actions:

(i) change the credit line;

(ii) change the annual percentage rate (APR);

(iii) change both the credit line and the annual percentage rate (APR);

or

(iv) making no changes.

24. The system according to claim 21, where the desired effect is associated with at least one of the bankcard accounts'"'"':

(i) net cash flow (NCF);

(ii) balance migration;

or

(iii) risk of credit losses.

25. The system according to claim 20, further comprising:

an analysis module for performing a regression tree analysis on the plurality of variables;

an establishment module for establishing the at least one behavior variable based on the regression tree analysis; and

a determination module for determining the plurality of interval breakpoints for each of the at least one behavior variable based on the regression tree analysis.

26. The system according to claim 20, further comprising:

a rule module for applying business rules to an output of the Markov Decision Process model; and

a generation module for generating a policy lookup table based on the business rules and the output of the Markov Decision Process model.

27. The system according to claim 20, further comprising a simplification module for simplifying the transition matrix based on one or more of the following techniques:

(i) creating a rectangular transition matrix with rows for the at least one control variable and the at least one behavior variable, and columns for the at least one behavior variable only;

(ii) combining at least one row of a transition matrix with an adjoining row, where the at least one row is characterized with low-frequency transitions;

(iii) combining at least one column of a transition matrix with an adjoining column, where the at least one column is characterized with low-frequency transitions; and

(iv) re-indexing a transition matrix to correctly capture all past and future account transitions.

28. The system according to claim 20, further comprising a simulation module for validating and supplementing the output of the Markov Decision Process model through a Monte Carlo simulation, where the Monte Carlo simulation is based at least partially on the account data.

29. A system for managing financial accounts, the system comprising:

means for collecting account data for a plurality of financial accounts, where the account data are associated with a plurality of variables;

means for establishing at least one control variable and at least one behavior variable based at least in part on the plurality of variables;

means for determining a plurality of interval breakpoints for the at least one control variable and the at least one behavior variable;

means for defining a plurality of account states based at least in part on the at least one control variable, the at least one behavior variable and the plurality of interval breakpoints;

means for forming a transition matrix based at least in part on the account data and the plurality of account states;

means for generating a Markov Decision Process (MDP) model based at least in part on the transition matrix; and

means for using the Markov Decision Process model to effect a change in the at least one control variable resulting in a desired effect on each of the plurality of account states.

30. The system according to claim 29, where the plurality of financial accounts are bankcard accounts and the at least one control variable comprises at least one of:

(i) credit lines;

or

(ii) annual percentage rates (APRs).

31. The system according to claim 29, where the account data comprise time-series data on at least one of:

(i) credit bureau variables;

(ii) credit bureau scores; and

(iii) bankcard account performance.

32. The system according to claim 30, where the change in the at least one control variable for the desired effect comprises at least one of the following actions:

(i) change the credit line;

(ii) change the annual percentage rate (APR);

(iii) change both the credit line and the annual percentage rate (APR);

or

(iv) making no changes.

33. The system according to claim 30, where the desired effect is associated with at least one of the bankcard accounts'"'"':

(i) net cash flow (NCF);

(ii) balance migration;

or

(iii) risk of credit losses.

34. The system according to claim 29, further comprising:

means for performing a regression tree analysis on the plurality of variables;

means for establishing the at least one behavior variable based on the regression tree analysis; and

means for determining the plurality of interval breakpoints for each of the at least one behavior variable based on the regression tree analysis.

35. The system according to claim 29, further comprising:

means for applying business rules to an output of the Markov Decision Process model; and

means for generating a policy lookup table based on the business rules and the output of the Markov Decision Process model.

36. The system according to claim 29, further comprising means for simplifying the transition matrix based on one or more of the following techniques:

(i) creating a rectangular transition matrix with rows for the at least one control variable and the at least one behavior variable, and columns for the at least one behavior variable only;

(ii) combining at least one row of a transition matrix with an adjoining row, where the at least one row is characterized with low-frequency transitions;

(iii) combining at least one column of a transition matrix with an adjoining column, where the at least one column is characterized with low-frequency transitions; and

(iv) re-indexing a transition matrix to correctly capture all past and future account transitions.

37. The system according to claim 29, further comprising means for validating and supplementing the output of the Markov Decision Process model through a Monte Carlo simulation, where the Monte Carlo simulation is based at least partially on the account data.

38. A method for managing financial accounts, where a Portfolio Control and Optimization model is represented as: