I got ML parameter estimates for the 3 parameters (lamda,R,sigma) of the MRW model (using the R code of Lovsletten and Rypdal, 2011).
Now I am at the point where I have to estimate volatility forecasts using the optimal linear prediction filter. I computed the auto-covariance function using the acf.mrw function and then used acf2AR() in order to derive the linear prediction filter. The output is a matrix, but I am unclear of how to derive the forecasts for the model from the matrix....any suggestions?