[R-SIG-Finance] rugarch NIG shape and skew parameters
alexios at 4dscape.com
Fri Jul 3 03:03:04 CEST 2015
(skew,shape) are the location-scale invariant parameters (\rho,\zeta) in
one of the parametrizations of the GH distribution (the '2nd
parametrization' in Prause 1999). They jointly map onto
(\alpha,\beta,\delta,\mu) - the standard parameters of the GH
distribution as shown in equations 70-82 of Section 2.3.5 of the vignette.
The (\bar \alpha,\bar \beta) you are referring to is a different
location-scale invariant parametrization (the '4th').
One of the nice features of the (\rho,\zeta) parametrization is that
you can immediately translate it to the (\xi,\chi) parametrization (the
'3rd') which asymptotically describes the skewness and kurtosis,
respectively, of the hyperbolic distribution (lambda=1). See
Barndorff-Nielsen et al 1985 for details.
However, for the particular limiting case of the GH Skew Student
(discussed seperately in Section 2.3.6), the \bar \beta
parametrization (skew) together with nu (shape) is used.
If you want to investigate the maximum domain of skewness-kurtosis of
the various distributions, have a look at the skdomain function, an
example of which is available on the 'Distributions Functions' section
of this rather old post:
On 02/07/2015 20:29, Andy Tang wrote:
> Hi Alex,
> In your rugarch package, the garch+nig model has coefficients of 'skew' and 'shape'. These two parameters map to alpha_bar/beta_bar, not alpha/beta in the NIG parameter set, right?
> Andy Tang, CFA
> Quantitative Analyst
> Morningstar, Inc.
> Morningstar. Illuminating investing worldwide.
> +1 312-384-4839 voice
> andy.tang at morningstar.com<mailto:andy.tang at morningstar.com>
> This e-mail contains privileged and confidential information and is intended only for the use of the person(s) named above. Any dissemination, distribution, or duplication of this communication without prior written consent from Morningstar is strictly prohibited. If you have received this message in error, please contact the sender immediately and delete the materials from any computer.
> [[alternative HTML version deleted]]
> R-SIG-Finance at r-project.org mailing list
> -- Subscriber-posting only. If you want to post, subscribe first.
> -- Also note that this is not the r-help list where general R questions should go.
More information about the R-SIG-Finance