System for computing probability distribution of loan losses

• US 7,050,999 B1
• Filed: 03/31/2000
• Issued: 05/23/2006
• Est. Priority Date: 04/22/1999
• Status: Expired due to Fees

• Current Assignee: Resona Holdings Incorporated
• Original Assignee: Resona Holdings Incorporated
• Inventors: Ota, Hiroyuki
• Primary Examiner(s): Kazimi, Hani M.
• Assistant Examiner(s): Subramanian, Narayanswamy
• Application Number:
  US09/540,890
• Time in Patent Office: 2,244 Days
• Field of Search: 705/70, 705/30, 705 35- 37, 705/42, 705/26
• US Class Current: 705/37
• CPC Class Codes: G06Q 10/0635 Risk analysis of enterprise or organisation activities
• CPC Class Codes: G06Q 30/0643 Graphical representation of items or shoppers
• 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 40/04 Trading; Exchange, e.g. stocks, commodities, derivatives or currency exchange

Abstract

A system for computing probability distribution of loan losses includes a loan amount and bankruptcy probability input device for inputting loan amounts and bankruptcy probabilities of individual loan customers, a characteristic function calculating device for calculating a characteristic function from these loan amounts and bankruptcy probabilities. A probability distribution calculating device calculates probability distribution by Fourier transform inversion of the characteristic function, and a probability distribution output device outputs the calculated probability distribution as a graph through a printer.

Claims

1. A computer based system for computing probability distribution of loan losses in a financing organization having a plurality of loan customers, comprising:

a scenario acquiring means for acquiring loan amounts of each loan customer and acquiring a plurality of bankruptcy probabilities of each said loan customer by predicting future fluctuations, and then using these values as a plurality of scenarios;

a characteristic function calculating means for calculating characteristic functions for each said scenario on the basis of said loan amounts and said bankruptcy probabilities acquired by said scenario acquiring means;

a probability distribution calculating means for calculating probability distributions for each said scenario by Fourier transform inversion of said characteristic functions calculated by said characteristic function calculating means;

an average probability distribution calculating means for calculating an average probability distribution which is the average of said probability distributions for each said scenario; and

a probability distribution output means for outputting said average probability distribution calculated by said average probability distribution calculating means,

wherein said scenario acquiring means expresses probabilities of bankruptcy of said loan customers by a function, and acquires said plurality of bankruptcy probabilities according to said function,

wherein said function expressing probabilities of said loan customers is

2. A computer based system for computing probability distribution of loan losses in a financing organization having a plurality of loan customers, comprising:

a scenario acquiring means for acquiring loan amounts of each loan customer and acquiring a plurality of bankruptcy probabilities of each said loan customer by predicting future fluctuations, and then using these values as a plurality of scenarios;

a characteristic function calculating means for calculating characteristic functions for each said scenario on the basis of said loan amounts and said bankruptcy probabilities acquired by said scenario acquiring means;

a probability distribution calculating means for calculating probability distributions for each said scenario by Fourier transform inversion of said characteristic functions calculated by said characteristic function calculating means;

an average probability distribution calculating means for calculating an average probability distribution which is the average of said probability distributions for each said scenario; and

a probability distribution output means for outputting said average probability distribution calculated by said average probability distribution calculating means,

wherein said scenario acquiring means expresses probabilities of bankruptcy of said loan customers by a function, and acquires said plurality of bankruptcy probabilities according to said function,

wherein said function expressing probabilities of said loan customers is

Norm(Y_k−

a_ku)

where k indicates each loan customer, u is a random variable, and a_k is a constant,

wherein Norm ( ) is a cumulative distribution function of a standard normal distribution and Y_k is a constant.

3. A computer based system for computing probability distribution of loan losses in a financing organization having N loan customers k=1 . . . N comprising:

an input means for inputting loan amounts M_k to each said N loan customers k=1 . . . N and bankruptcy probabilities p_k thereof;

a loan customer calculating means for calculating the number N of said loan customers on the basis of said loan amounts M_k and/or said bankruptcy probabilities p_k inputted by said input means;

a characteristic function calculating means for calculating a characteristic function

4. The system for computing probability distribution according to claim 3 wherein 2²ⁿ−

1 which is the greatest value of m in said characteristic function calculating means is a value not less than the sum of all loan amounts.

5. The system for computing probability distribution according to claim 3 wherein 2²ⁿ−

1 which is the greatest value of m in said characteristic function calculating means is a value not less than the minimum of values for which probability of loan losses can be regarded substantially zero in the calculation process.

6. The system for computing probability distribution according to claim 3 further comprising a loan amount rounding means for rounding each said loan amount inputted through said input means to a number integer times of a predetermined unit, and

2²ⁿ−

1 which is the greatest value of m in said characteristic function calculating means being a value not less than the quotient obtained by dividing the sum or all loan amounts by said predetermined unit.

7. The system for computing probability distribution according to claim 3 further comprising a loan amount rounding means for rounding each said loan amount inputted through said input unit to a number integer times of a predetermined unit, and

2²ⁿ−

1 which is the greatest value of m in said characteristic function calculating means being a value not less than the quotient obtained by dividing by said predetermined unit the minimum of values for which probability of loan losses can be regarded substantially zero in the calculation process.