System and method for evaluating residual values of products

• US 20030074296A1
• Filed: 08/28/2002
• Published: 04/17/2003
• Est. Priority Date: 08/30/2001
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Blanchard, Jean Louis, Messatfa, Hammou, Lorin, Stephane, Pavillon, Christelle
• Granted Patent: US 7,546,273 B2
• Time in Patent Office: Days
• US Class Current: 705/36
• CPC Class Codes: G06Q 20/102 Bill distribution or payments
• 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 40/06 Asset management; Financial planning or analysis

Abstract

A method and system to evaluate residual value of leased products, comprising a learning phase and an application phase. Each phase has two steps. The first consists in analyzing lease or product main parameter while the second uses other lease conditions or product characteristics. Analysis is based on a grid wherein several specific points are determined during the learning phase. An extrapolation of these points according to the parameters during the application phase gives a first approximation of residual value, which is adjusted using lease conditions or product characteristics. During the learning phase, the system determines the relevant lease conditions or product characteristics and the associated corrections used during the application phase.

Claims

1. A method to learn residual value factor behavior of leased products, using a learning database containing lease conditions and product characteristics and resale value factors associated therewith, comprising the steps of:

selecting at least one lease condition or product characteristic of said learning database;

specifying values of said at least one lease condition or product characteristic, determining residual value factors associated with said specified values of said at least one lease condition or product characteristic from said learning database and building a first table storing said specified values of said at least one lease condition or product characteristic and said residual value factors associated therewith;

estimating first residual value factors for records of said learning database from said first table; and

,

selecting at least one other lease condition or product characteristic of said learning database, determining residual value factor corrections associated with modalities of said at least one other lease condition or product characteristic from said learning database and said first residual value factors, and building at least one other table storing the modalities of said at least one other lease condition or product characteristic and said residual value factor corrections associated therewith;

wherein said first and at least one other table characterize the residual value factor behavior of the products contained in said learning database.

2. The method of claim 1, wherein said steps of determining residual value factors and building said first table comprise the steps of:

setting a window value for each of said at least one lease condition or product characteristic;

for each of said specified values of said at least one lease condition or product characteristic;

storing said specified value in said first table;

selecting records of a learning database whose values according to said at least one lease condition or product characteristic belong to a range defined by said specified value and said window values;

extrapolating a residual value factor from the residual value factors associated with said selected records; and

,

storing said extrapolated residual value factor in said first table in association with said specified value.

3. The method of claim 2, wherein said extrapolated residual value factor is a median value of the residual value factors associated with said selected records.

4. The method of claim 1, wherein said step of estimating first residual value factors comprises the steps of:

for each record of said learning database;

selecting specified values of said first table bordering the value of said record corresponding to said at least one lease condition or product characteristic; and

,

extrapolating a first residual value factor from the residual value factors associated to said selected specified values of said first table.

5. The method of claim 4, wherein said extrapolation is a linear extrapolation.

6. The method of claim 1, wherein said step of estimating first residual value factors comprises the steps of:

for each record of said learning database,

comparing the value of said record corresponding to said at least one lease condition or product characteristic with said specified values of said first table; and

,

if said value of said record corresponding to said at least one lease condition or product characteristic is bordered on specified values of said first table, determining a first residual value factor of said record by performing a linear extrapolation on residual value factors associated with specified values of said first table bordering said value of said record corresponding to said at least one lease condition or product characteristic,

else, determining a first residual value factor of said record by performing a linear extrapolation on residual value factors associated with specified values of said first table bordering said value of said record corresponding to said at least one lease condition or product characteristic wherein first residual value factors of missing bordering specified values of said first table are estimated using a global residual value factor variation.

7. The method of claim 1, wherein said step of selecting at least one other lease condition or product characteristic of said learning database, determining residual value factor corrections associated with the modalities of said at least one other lease condition or product characteristic from said learning database and said first residual value factors and building at least one other table storing the modalities of said at least one other lease condition or product characteristic and said residual value factor corrections associated therewith, comprises the steps of:

determining an error criterion to evaluate lease condition or product characteristic influence on residual value factor correction;

setting an error variable to the greatest reachable error value;

selecting a first variable to be tested;

selecting a modality of said selected first variable to be tested and the learning database records that correspond to said selected modality;

determining a residual value factor correction according to a difference between first residual value factor estimations and resale value factors of said selected records corresponding to said selected modality;

repeating the preceding two steps until all modalities of said selected variable to be tested have been selected;

estimating the error associated with said selected variable;

if said estimated error is lower than said error variable;

setting said error variable to said estimated error;

memorizing the reference of said selected variable;

storing modalities of said selected variable and associated correction in said at least one other table; and

,

repeating the above steps until all the variables to be tested have been selected.

8. The method of claim 7, wherein said determined residual value factor correction is a median value of the differences between first residual value factor estimations and resale value factors of said selected records corresponding to said selected modality.

9. The method of claim 1, further comprises a step analyzing said learning database records to detect aberrant records, and isolating the aberrant records from said learning database.

10. The method of claim 1, wherein said specified values of said at least one lease condition or product characteristic are determined so that learning database records are uniformly distributed in said first table.

11. The method of claim 1, further comprising the steps of:

inputting a value of said leased product corresponding to said at least one lease condition or product characteristic;

determining a first residual value factor estimation from said first table;

inputting the value of said leased product corresponding to said at least one other lease condition or product characteristic;

determining at least one residual value factor correction from said at least one other table; and

,

determining the residual value factor of said leased product by adding said first residual value factor evaluation with said at least one residual value factor correction.

12. The method of claim 11, wherein said step of determining a first residual value factor from said first table, comprises the steps of:

selecting specified values of said first table bordering the value of the leased product corresponding to said at least one lease condition or product characteristic; and

,

extrapolating a first residual value factor from the residual value factors associated with said selected specified values of said first table.

13. The method of claim 12, wherein said extrapolation is a linear extrapolation.

14. The method of claim 11, wherein said step of determining a first residual value factor from said first table, comprises the steps of:

comparing the value of said leased product corresponding to said at least one lease condition or product characteristic with said specified values of first table; and

,

if said value of said leased product corresponding to said at least one lease condition or product characteristic borders specified values of said first table, determining a first residual value factor of said leased product by performing a linear extrapolation on residual value factors associated with specified values of said first table bordering said value of said leased product corresponding to said at least one lease condition or product characteristic;

else, determining a first residual value factor of said leased product by performing a linear extrapolation on residual value factors associated with specified values of said first table bordering said value of said leased product corresponding to said at least one lease condition or product characteristic wherein residual value factors of missing bordering specified values of said first table are estimated using a global residual value factor variation.

15. The method of claim 1, wherein said at least one lease condition or product characteristic comprises a lease duration.

16. The method of claim 1, wherein said leased products are vehicles.

17. The method of claim 16, wherein said at least one lease condition or product characteristic comprises a lease mileage.

18. A computerized system comprising means adapted to implement the method of claim 1.

19. A program storage device readable by machine, tangibly embodying a program of instructions executable by the machine to perform steps of the method as described in claim 1.