Hi,

I am trying to write code for the rolling windows forecasting using HAR-RV model.

My codes are:

I am not sure whether my forecasted values are rolling windows forecasts or not. I can see on the script I succeeded re-estimated rolling windows regressions (that s why i did not put -q there). But, I am not sure my one-step-ahead forecasted values are generated with sliding re-estimated regressions. [I am not sure should I change my 5th line with fcast 336+i 336+i 1 forecasted --rolling] Or what should I do to get one-step-ahead rolling windows forecasts in this loop command?

Does anyone know about that. I think I am missing something here on codes. I am sorry I am making busy this line about this a couple of times, but I am trying to understand and there is no one i can ask about that except here.

Thank you very much.