garchfilter: GARCH filter applied separately to log returns

In YafeiXu/CopulaModel: CopulaModel: Dependence Modeling with Copulas

Description

GARCH filter applied separately to a matrix of log returns, the d log returns should have the same dates, a subset of rows can be chosen

Usage

gfiltersubset(lgret,ar,m1,m2,iprint=F)

Arguments

lgret	Nxd matrix of log returns of d financial assets
ar	TRUE for univariate GARCH(1,1)-AR(1) for model, FALSE for GARCH(1,1) model
m1	first row of lgret to use for the subset
m2	last row of lgret to use for the subset
iprint	print flag for GARCH estimates, F by default

Details

This function requires library fGarch and its dependent libraries.

Value

filter	GARCH filtered data (nxd, where n=m2-m1+1)
uscore	empirical uniform scores (nxd)
zscore	empirical normal scores (nxd)
uscmodel	model-based uniform scores (nxd)
zscmodel	model-based normal scores (nxd)
sigmat	matrix of estimated volatilities (nxd)
coef	matrix of GARCH parameters (6xd or 5xd for ar=T or ar=F respectively), the parameters are mu, (ar1), omega, alpha1, beta1, shape.

Examples

## Not run: 
data(euro07lr) # euro07lr data set is 239x7
library(fGarch)
out=gfiltersubset(euro07lr,ar=FALSE,1,200,iprint=TRUE) 
## End(Not run)

YafeiXu/CopulaModel documentation built on May 9, 2019, 11:07 p.m.