Multivariate Distribution = Marginals + Copulas
The Copula Marginal Algorithm (CMA) is a simple two step recipe to manipulate multivariate distributions under Fully Flexible Probabilities.
CMA can quickly decompose any multivariate distribution between unique (marginals) and their shared components (copulas). This approach can add a high level of flexibility for estimation and simulation purposes.
Install the development version of CMA from github with:
# install.packages("devtools")
devtools::install_github("Reckziegel/CMA")-
Meucci, Attilio, A New Breed of Copulas for Risk and Portfolio Management (May 22, 2011). Risk, Vol. 24, No. 9, pp. 122-126, 2011, Available at SSRN: https://www.ssrn.com/abstract=1752702
-
Meucci, Attilio, A Short, Comprehensive, Practical Guide to Copulas (May 20, 2011). GARP Risk Professional, p. 22-27, October 2011, Available at SSRN: https://www.ssrn.com/abstract=1847864 or http://dx.doi.org/10.2139/ssrn.1847864
-
Attilio Meucci (2021). Copula-Marginal Algorithm (CMA) (https://www.mathworks.com/matlabcentral/fileexchange/32701-copula-marginal-algorithm-cma), MATLAB Central File Exchange. Retrieved September 23, 2021.