[R-SIG-Finance] Backtesting

[alexios ghalanos]

What one can do is take more care when submitting what is meant to be 
a reproducible example: You are passing an undefined "spec3" when in
fact you want to pass "spec" as defined by yourself below (which when I
try it works just fine and returns the eGARCH-sstd rolling model).

Regards,
Alexios

On Thu, 2012-01-19 at 13:02 +0000, Papa Senyo wrote:
> Dear ALL.
> Please, find the code below;
> spec = ugarchspec(variance.model = list(model = "eGARCH"),
> distribution.model ="sstd")
>  roll = ugarchroll(spec3, data = xx, n.ahead = 1, forecast.length =
> 875,
>  refit.every = 100, refit.window = "recursive", parallel = TRUE,
>  parallel.control = list(pkg = "snowfall", cores = 10), solver =
> "solnp",
>  solver.control = list(tol = 1e-05, delta = 1e-06, trace = 0),
>  calculate.VaR = TRUE, VaR.alpha = c(0.01, 0.05))
>  report(roll, type = "VaR", n.ahead = 1, VaR.alpha = 0.01, conf.level
> = 0.99)
> xx is a univariate time series.
> The problem is the even though the distribution.model="sstd" it
> reports EGARCH-norm
> and also changing the backtest length  have no effect.
> Also when for the VaR report both the conditional average and the
> unconditional average are rejected, what esle can one do and what does
> it imply.
> Kind regards
> Papa
>