[R-SIG-Finance] Query about mcsGARCH (rugarch package)

alexios ghalanos alexios at 4dscape.com
Tue Jun 24 14:27:02 CEST 2014


Hi Will,

The model presented in the blog used the mean function to calculate the
diurnal component, but I had since (and noted that in the post) changed
the mean to the median. It was perhaps not a good idea to do so without
allowing for the choice to be set by the user. For the time being, I've
reversed the change to use the mean again until such time as I provide
an option for choosing between the two. You can download the latest
revision from my bitbucket repository (see the downloads section of blog).

Cheers,

Alexios


On 24/06/2014 10:23, Ferguson, William wrote:
> Hi,
> 
> I have looked at the article on the multiplicative component GARCH model found at http://unstarched.net/2013/03/20/high-frequency-garch-the-multiplicative-component-garch-mcsgarch-model/, and would like to adapt the example given for FX data.
> 
> I am trying to reproduce the initial plots that show the four volatility components of the model (second figure), and only the plot of Daily-Forecast is the same. My Diurnal plot seems to be lower than that in the example, which presumably then affects the Stochastic and Total plots.
> 
> I think that both data sets used are fine as the ACF plot and Daily_Forecast plot are both the same as in the example. I assume therefore that the problem lies in the following lines:
> 
> spec = ugarchspec(mean.model = list(armaOrder = c(1, 1), include.mean = TRUE), variance.model = list(model = 'mcsGARCH'), distribution = 'nig')
> fit = ugarchfit(data = R_i, spec = spec, DailyVar = f_sigma^2)
> 
> Is anyone able to explain the discrepancy? I wonder whether perhaps different versions have different function defaults, or something like that, though I can't spot a problem. I am using R version 3.1.0 and rugarch version 1.3-1.
> 
> Many thanks!
> 
> Will
> 
> 	[[alternative HTML version deleted]]
> 
> _______________________________________________
> R-SIG-Finance at 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