Let's say the data input is called
data, which is in the form of a financial time series that will be fitted into an ARMA-GARCH model. The GARCH(1,1) is used because it is the best fit for a financial time series with volatility and heteroscedasticity.
ugarchspec (comparing the BIC values) I found that the best fit is ARMA(2,1)GARCH(1,1) with parameters
mu = -0.322925157,ar1= 1.904752272, ar2= -0.918770379, ma1=-0.924142232,
omega= 0.004355971, alpha1= 0.198723766,shape= 6.520216307, beta1 =0.713514758
Another method of fitting was to look at the acf and pacf of the data and instinctively find a good match, which is ARMA(1,0)GARCH(1,1) with parameters
mu = -0.453529886,ar1= 0.933776546, omega= 0.006608471, alpha1= 0.257053963,shape= 6.533376243, beta1 =0.622149898
I'm wondering how can I achieve the residuals from fitting
data to each of these models and parameters. I tried using
ugarchpath to input the parameters and each time I generate it comes up with a different model.
So, what I'm asking is what code or function can I use to find the residual manually? Because with the function
resid, it gives the result from the first model ARMA(2,1)GARCH(1,1) because of the lowest BIC value. How can I get it from the ARMA(1,0)GARCH(1,1)?