I have the following garch code, I am modeling 10 years of data for the SP500
model33 <- ugarchspec(variance.model = list(model="sGARCH",
garchOrder=c(2, 1)),
mean.model = list(armaOrder=c(1, 1)),
distribution.model = "norm")
mod33 <- ugarchfit(spec=model33,data=difsfts)
forc1 = ugarchforecast(mod33, n.ahead=20)
However my output is showing
Model: sGARCH
Horizon: 20
Roll Steps: 0
Out of Sample: 0
0-roll forecast [T0=1977-07-28 20:00:00]:
Series Sigma
T+1 -0.68565 13.78
T+2 -0.61457 14.23
T+3 -0.54608 14.37
T+4 -0.48009 14.52
T+5 -0.41650 14.66
T+6 -0.35523 14.80
T+7 -0.29619 14.92
T+8 -0.23931 15.04
T+9 -0.18450 15.16
T+10 -0.13168 15.26
T+11 -0.08079 15.36
T+12 -0.03176 15.46
T+13 0.01549 15.55
T+14 0.06101 15.64
T+15 0.10488 15.72
T+16 0.14714 15.80
T+17 0.18787 15.87
T+18 0.22711 15.94
T+19 0.26492 16.01
T+20 0.30136 16.07
Now my data is from 2009-2019 so I am not sure where it is picking up the date from 1977, my first point is 01-02-09 and my last is 12-30-19, also the forecasted values seem to be very low compared to what the previous values of "difsfts" (aka I differenced my dataset), which had closing values going from ranging as high as 40. I am not sure if this is a result of the forecasting command using the wrong date or a fault in the model itself.
plot(forc1,which="all")
This plot shows the forecasted values of the SP500 to be moving in a horizontal line more or less which is not accurate.