[R-sig-finance] Random Numbers

Kris kriskumar at earthlink.net
Fri Nov 18 17:10:17 CET 2005

 I dont quite follow what you mean? People do resampled eff frontier with bootstrapping/bootstrapping+jackknife but this is done on the correlation/covarianceestimation process.
If all you need is correlated rng take a look at V&R's MASS package rmvnorm in particular. alternatively you can use rnorm with chol to get the correlated RNG.

L.Isella wrote:

>Dear All,
>I would like to be able to generate long sequences of random numbers which I need to work out the resampled efficient frontier of a portfolio.
>People normally take random draws from some multivariate (say Gaussian, to keep it simple) distribution to simulate the returns of the various assets and generate many random portfolios.
>I accomplish that working with rnorm().
>However, this way, in the limit of very long sequences, the simulated asset returns have the chosen mean and std, but each sequence is totally uncorrelated from the other.
>Is there (in R) a way to produce sequences of Gaussian distributed random numbers with a chosen correlation? Regardless whether this is the proper way of calculating the resampled efficient frontier, I wonder it this has been implemented somewhere for R.
>R-sig-finance at stat.math.ethz.ch mailing list

R-sig-finance at stat.math.ethz.ch mailing list

More information about the R-sig-finance mailing list