good morning, I need some help.
I am trying to get Eviews to do the following:
I have estimated my regression using GARCH with data from 2006 until 2017.
to prove the existence of the day of the week effect I am forecasting the returns from 2017 January until 2017 December to compare the average errors of the forecast with the actuals.
to make this exercise fair, I need to update the estimation of the coefficients every new day that is forecast
I mean:
to forecast January 1 2017 I used the estimates up until 31 December 2016
to forecast January 2 2017 I used the estimates up until 1 January 2017
and so on.
at the beginning I thought I could use the static forecast tool, but then I realise that I the coefficients would be using the data until December 2017 to be calculated, and that is not efficient or accurate.
to do this I need to change the parameters of the estimation to include the next day each time,
my specific question is:
is t there a way I can do this programing Eviews to do it automatically and not having to do it one by one.
bare in mind that I have to do it 252 times minimum to complete a hole year of comparison.
thank you
how to forecast GARCH updating the estimated values in each time
Moderators: EViews Gareth, EViews Moderator
-
- Fe ddaethom, fe welon, fe amcangyfrifon
- Posts: 13319
- Joined: Tue Sep 16, 2008 5:38 pm
Re: how to forecast GARCH updating the estimated values in each time
thank you Gareth for the link, I have read it and still quit confuse
I have my regression as follows
rsp500 c rsp500(-1) rsp500(-2) rsp500(-3) rsp500(-4) .........(lagged regression of daily returns)
my sample size is from 1/01/2006 to 12/31/2017
I want to forecast al the returns for the hole 2017 year, with a daily roll over to correct the coefficients each day
what I mean is that when I forecast 1/01/2017 my estimates coefficients are calculated only using the actual data up until 12/31/2016, for the next day forecast 1/02/2017, my estimates coefficients are calculated using the actual date up until 1/01/2017 and so on.
can you please help me how to do the programing?
I need this for my dissertation, and I am not really good at programming.
thank you.
I have my regression as follows
rsp500 c rsp500(-1) rsp500(-2) rsp500(-3) rsp500(-4) .........(lagged regression of daily returns)
my sample size is from 1/01/2006 to 12/31/2017
I want to forecast al the returns for the hole 2017 year, with a daily roll over to correct the coefficients each day
what I mean is that when I forecast 1/01/2017 my estimates coefficients are calculated only using the actual data up until 12/31/2016, for the next day forecast 1/02/2017, my estimates coefficients are calculated using the actual date up until 1/01/2017 and so on.
can you please help me how to do the programing?
I need this for my dissertation, and I am not really good at programming.
thank you.
Re: how to forecast GARCH updating the estimated values in each time
this is what I am doing manually,
as you see in the file I have to change the sample in the equation estimation each time to include one more day ( the day after the one I am forecasting)
and manually change the forecast sample as well to get the new forecast
can Eviews do this automatically?
please tell me how
as you see in the file I have to change the sample in the equation estimation each time to include one more day ( the day after the one I am forecasting)
and manually change the forecast sample as well to get the new forecast
can Eviews do this automatically?
please tell me how
- Attachments
-
- forecast.docx
- (46.77 KiB) Downloaded 234 times
Who is online
Users browsing this forum: No registered users and 43 guests