AnalyzeTS (version 2.0)

forecastGARCH: Forecast GARCH Model

Description

This function get two object from 'Arima' class and 'garch' class, and then calculate to return forecasting answer of mean and variance of next day.

Usage

forecastGARCH(fitARMA, fitGARCH, r = 3, trace = FALSE, newxreg = NULL)

Arguments

fitARMA

A object from 'Arima' class.

fitGARCH

A object from 'garch' class.

r

Rounds the answer to the specified number of decimal places (default 3).

trace

Logical. Trace optimizer output?

newxreg

A covariates value of next day for ARMAX-GARCH mdels.

Value

ARCH

GARCH coefficients.

ARMA

ARMA coefficients.

forecast

Forecasting answer:Point: forecasting time.res: forecasting residual.res^2: res square.SSL.forecast: forecating mean value.VAR.forecast: forecasting variance value.

Examples

#Load data
library(TTR)
data(ttrc)

#Calculate SSL series
t<-ts(ttrc[,"Close"],start=1,frequency=5)
ln.t<-log(t)
r<-diff(ln.t)

#Find a ARIMA model
fit1<-arima(r,order=c(4,0,0))

#Find a GARCH model
res1<-resid(fit1)
library(tseries)
fit2<-garch(res1,order=c(2,1),trace=0)

#Forecasting
forecastGARCH(fit1,fit2,r=6,trace=TRUE)
forecastGARCH(fit1,fit2,r=6)