05-28-2014 12:04 PM
Does anybody know how to use "errormodel y~general()" in Proc Model to specify and maximize your own likelihood function? Any comment is appreciated. I am trying to implement Maheu and McCurdy (2004) Journal of Finance paper, the GARJI model. The model involves a poisson jump process and I need to specify my own likelihood function. I think the first step is to estimate GARCH(1,1) model without using h.var statement and specify my likelihood function. Thank you in advance.