Simulations with Exact Means and Covariances
Attilio Meucci
Bloomberg ALPHA, Portfolio Analytics and Risk
June 7, 2009
Abstract:
We present a simple method to generate scenarios from multivariate elliptical distributions where the sample mean and covariances match the respective population moments. This methodology easily applies to large numbers of scenarios and large-dimensional distributions. We show an application to the risk management of a book of options.
Keywords: matrix Riccati equation, antithetic variables, affine equivariance, affine transformations, copula-marginal factorization, correlation stress-testing
JEL Classifications: C1, G11
For the paper, pls see
For the code, pls see
Monday, August 10, 2009
Subscribe to:
Post Comments (Atom)
Posts
No comments:
Post a Comment