I am reading this article which is talking about a GARCH trading strategy. I follow the steps and tried different parameters like window size (stock historical data). From the back test, I can see that different window size can affect the result.
I would like to ask that if there is a way to compare different models so that I can choose a better one?
I have read another thread that is talking about AIC as a selection criteria. I can also see that there is a log likelihood from R. But I am afraid I cannot compare them directly between these two models.
Here are graphs of log returns of N225, with different window size.
