[R-SIG-Finance] rugarch package

[alexios ghalanos]

Hi Jun,

This is likely related to the time format of the data you are passing to 
the ugarchfit function ('data' argument). Ideally it should be an xts 
object, with a POSIXct (or Date) index, and you have also set your 
timezone (e.g. Sys.setenv(TZ="GMT") ).

You should really take some time to familiarize yourself with R's 
Date/Time functions and the xts package (it will payoff in the long run, 
IMHO).

Best,

Alexios

On 23/06/2013 17:38, jun wang wrote:
> Hi, Alexios,
>
> I am using the rugarch package to get the conditional variance of the
> uni variate garch model. But the result i got seems a little weird, see
> below:
>
> uspec = ugarchspec(mean.model = list(armaOrder = c(0,0),include.mean=TRUE),
>                     variance.model = list(garchOrder = c(1,1), model =
> "sGARCH"),
>                     distribution.model = "ghyp")
> fit1=ugarchfit(uspec,data=jpm)
> fit2=ugarchfit(uspec,data=sp500)
>
> a=sigma(fit1)
> b=sigma(fit2)
> H=cbind(1/a,1/b)
>
> In the matrix H, i had the following format:
> 1970-01-01 19:00:00 0.10648731 0.004840093
> 1970-01-02 19:00:00 0.13220160 0.003223305
>
> Any way to get rid of  1970-01-01 19:00:00?
>
> Many thanks!!!
>
> Jun
>