[R-SIG-Finance] rugarch roll forecast

Alexios Ghalanos @lexio@ @ending from 4d@c@pe@com
Mon Dec 3 18:22:29 CET 2018


The forecasts are for y[t] given information up to [t-1], and aligned to realized at y[t] for comparison.

Alexios

> On Dec 3, 2018, at 7:15 AM, Владимир Иванов <vivanov879 using ya.ru> wrote:
> 
> Hey, Alexios and fellow R users.
>  
> A small thing to clarify. When I run the
>  
> roll = ugarchroll(spec, xts(df$dprice, as.POSIXct(df$time)), forecast.length = 100, refit.every = 25)
>  
> and then analyze the roll using forecast$density table. I get 100 forecasts.
>  
>  
> > tail(density)
> ====================
>                             Mu    Sigma    Skew    Shape Shape(GIG)  Realized
> 2018-11-30 09:46:10 -11.717360 11.25498 1.00654 7.013685          0 -16.85714
> 2018-11-30 09:46:15   2.521165 10.94367 1.00654 7.013685          0  11.60714
> 2018-11-30 09:46:20   1.032400 11.25478 1.00654 7.013685          0  22.91667
>  
> ====================
>  
> It produces forecasts for Y{t} given all information available up to time t.
> That is, for, example, the 2018-11-30 09:46:15 forecast is made given information of up until and including 2018-11-30 09:46:10.
>                                   the 2018-11-30 09:46:20 forecast is made given information of up until and including 2018-11-30 09:46:15.
> And the model is refitted every 25 steps, that is, since the time step here is 5 seconds, every 5 seconds X 25 = 125 seconds
>  
> Is my thinking correct?
> _______________________________________________
> R-SIG-Finance using r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-sig-finance
> -- 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 mailing list