[R-SIG-Finance] ugarchroll with moving window - failed to invert hessian

Michal Maganlal m|ch@|hm @end|ng |rom gm@||@com
Thu Apr 25 17:58:20 CEST 2019


I'm currently finishing my master thesis in Forecasting with ARMA-GARCH
models but I'm having some trouble forecasting with moving windows using
ugarchroll.

I'm trying to run the following code:

# GARCH(1,1) - t-student
specgarch11n = ugarchspec(mean.model = list(armaOrder = c(0,0),
include.mean = TRUE, archm=TRUE, archpow=2),
                          variance.model = list(model = "sGARCH",
garchOrder=c(1,1)),
                          distribution.model = "std")
specgarch11n
fitgarch11n <- ugarchfit(specgarch11n, data = returns_eurusd)
fitgarch11n

 #---Moving Window Forecast---
 forecast1 <- ugarchroll(specgarch11n, n.ahead = 1, forecast.length = 1507,
refit.window = "moving", window.size = 250, refit.every = 250,
                        data = returns_eurusd, solver = "hybrid")
 View(forecast1 using forecast$density$Sigma)

After running it I get the following error:

Warning messages:
1: In .makefitmodel(garchmodel = "sGARCH", f = .sgarchLLH, T = T, m = m,  :
  rugarch-->warning: failed to invert hessian

2: In .makefitmodel(garchmodel = "sGARCH", f = .sgarchLLH, T = T, m = m,  :
  rugarch-->warning: failed to invert hessian

3: In .makefitmodel(garchmodel = "sGARCH", f = .sgarchLLH, T = T, m = m,  :
  rugarch-->warning: failed to invert hessian

4: In .rollfdensity(spec = spec, data = data, n.ahead = n.ahead,
forecast.length = forecast.length,  :
  non-converged estimation windows present...resubsmit object with
different solver parameters...

I have tried to follow Alexios suggestions on this post but nothing seems
to work:
http://r.789695.n4.nabble.com/rugarch-package-quot-Warning-Message-quot-for-GARCH-Normal-td4643515.html

Can someone please help me solving this issue?

Kind regards,
Michal

	[[alternative HTML version deleted]]



More information about the R-SIG-Finance mailing list