[R-SIG-Finance] rugarch n.ahead forecasts
Eliano Marques
eliano.m.marques at gmail.com
Tue Sep 29 16:40:39 CEST 2015
Community,
I have a quick question that perhaps is easy to solve regarding n.ahead forecasts.
So for example:
variance_model = list(model = "sGARCH", garchOrder = c(1, 1),
submodel = NULL,
external.regressors = NULL,#as.matrix(Regressors), #External Regressors will include the news and tweets and market returns
variance.targeting = FALSE)
mean_model = list(armaOrder = c(1,0), include.mean = TRUE, archm = FALSE,
archpow = 1, arfima = FALSE,
external.regressors = as.matrix(Regressors),
archex = FALSE)
distribution_model = "norm"
spec = ugarchspec(mean.model = mean_model, variance.model = variance_model, distribution = distribution_model)
fit2 = ugarchfit(data = Returns, solver = "gosolnp", spec = spec,out.sample = 66 , DailyVar = f_sigma^2,
solver.control=list(n.restarts=5))
Say now I want to forecast ahead the next two days (66 - intraday data):
forc = ugarchforecast(fit2, n.ahead = 66, n.roll = 0) # Forecasting function
After researching in the documentation and other posts I just have a few doubts:
1 - Because the fit function has a out.sample that = n.ahead forecasts do I still need to provide the list of external regressors inside the forecast function?
2 - The >2 n.ahead forecast are calculated using the previous forecast of the mean and variance equation? I’m assuming yes given what is written in the documentation but better to confirm.
Thank you for your help,
Eliano
More information about the R-SIG-Finance
mailing list