System and method for performing risk analysis

1. A computerized data processing system including a computer for performing risk analysis of a portfolio, the system comprising:

a modeling and calibration unit implemented on the computer, configured to describe d risk factors as random variables X 1 to X d , the random variables being related to each other by a correlation matrix ρ, to form m groups of the random variables X 1 to X d , to describe the random variables X 1 to X d as a d-dimensional random vector X, forming m subvectors, each subvector including one group of the random variables X 1 to X d , and to describe dependencies of the risk factors as the implicit copula of a d-dimensional random vector Y, the random vector Y including m subvectors Y k (k=1 to m, m≠1, d, m, and k are natural positive numbers), each subvector Y k has a t-distribution with a parameter ν k describing a degree of freedom, and a copula of each subvector Y k is a t-copula;

an input unit implemented on the computer, configured to enter or choose calibration data to obtain, by using the modeling and calibration unit, values for the ν k describing the degree of freedom for each of the m subvectors Y k separately and to obtain values for a correlation matrix ρ for the random variables X 1 to X d , to enter or choose at least one risk mapping function L(X), and to enter portfolio data of the portfolio to be analyzed;

a simulation unit implemented on the computer, configured to simulating realization of the d risk factors by using the correlation matrix ρ, the parameters ν k describing the degrees of freedom, the at least one risk mapping function L(X), and portfolio data of the portfolio; and

an output unit implemented on the computer, configured to generate output data resulting form the simulation unit in a form of at least one of a risk measure or a price.

2. The computerized system according to claim 1 , wherein the

input unit is further configured to enter at least three input levels, a first level including a first input unit configured to enter or choose the calibration data, a second level including a second input unit configured to enter or choose the least one risk mapping function L(X), and a third level with third input unit configured to enter the portfolio data.

3. The computerized system according to claim 1 , further comprising a data storage unit configured to store the historical data.

4. The computerized system according to claim 1 , wherein the

input unit is further configured to group the d risk factors manually or configured to manually choose a grouping from a range of several kind of groupings.

5. The computerized system according to claim 1 , further comprising:

a storing unit in the computer configured to store the correlation matrix ρ and the parameters ν k describing the degrees of freedom.

6. A computer implemented method for performing risk analysis of a portfolio on a computer by combining d interdependent risk factors to determine a risk measure or a price, the method comprising:

a step of building a model with the computer by

describing the d risk factors as random variables X 1 to X d being related to each other by a correlation matrix ρ,

forming m groups of the random variables X 1 to X d ,

describing the random variables X 1 to X d as a d-dimensional random vector X to form m subvectors, each subvector including one group of the random variables X 1 to X d , and

describing dependencies of the d risk factors as an implicit copula of a d-dimensional random vector Y, the random vector Y including m subvectors Y k (k=1 to m), wherein each subvector Y k has a t-distribution with unknown values of parameters V k , the values of parameters V k describing degrees of freedom, and wherein a copula of the values of parameters V k being a t-copula;

a step of calibrating the model with the computer of said step of building by using historical data to obtain values for the parameters V k , describing degrees of freedom for each of the m subvectors Y k separately, and to obtain values for the correlation matrix ρ for all the random variables X 1 to X d , wherein d, m and k are natural positive numbers;

a step of simulating realization of the d risk factors with the computer by using the calibrated model of the step of calibrating; and

a step of generating output data with the computer in form of at least one of a risk measure or price based on said step of simulating.

7. The method according to claim 6 , wherein d is equal or greater than 4 and each group of the random variables X 1 to X d includes at least two random variables.

8. The method according to claim 6 , wherein in said step of simulating on the computer, Kendal's tau is estimated for every group of subvectors Y k .

9. The method according to claim 6 , wherein said step of calibrating the model with the computer further comprises a maximum likelihood estimation of the m degrees of freedom.

10. The method according to claim 6 , wherein the random variables X 1 to X d are grouped in m groups according to predefined aspects, countries, or industrial sectors.

11. The method according to claim 6 , wherein said step of building a module with the computer further comprises building the model by specifying normally distributed marginals.