VAR(2) BEKK MGARCH t=3

[amcqeen]

Hallo, following my previous post a few day ago you said when i have a VAR(2) i should start at least from t=3
But my data are in excel and the first row are the names of the series so in the code i have "1 1025" and then i have gstart t=3
Because the first row are the series names should i have started from t=4 ?

Similarly when i had VAR(1) i had gstart at t=2.

Many thanks in advance
Maria

[amcqeen]

[color=#0000FF]"
open data arg_W1C_RD.xls
data(format=xls,org=columns) 1 1025 lmr fxr fxr2 mmrus mmrjp nar twr twr12 rdyj rdyf rdmco rdmca rdmfo lmrrd fxr2rd mmrusrd narrd twrrd
*
*
compute gstart=3,gend=1025"[/color]

My data have 1024 observations and i added 1-1025 because the excel has the names on row 1.
Then when i say gstart at t=3 for VAR(2) does it start from data observation 3 and considers i wrongly put 1025 data points since there are only 1024 in both points?



Thank you!

[TomDoan]

The row of labels doesn't count as an observation. If you have 1024 rows of numbers, you have 1024 entries. If you're running a VAR(2), then you can't start earlier than 3. If, in addition, you have to transform into returns within RATS, you will lose another data point doing that. If the data on the Excel file are usable as is, then you can start at 3.

[amcqeen]

Ignore the question, i have found the answer!
I used view and saw the data that Rats read, it reads as t=1 my first data observation and t=1024 my last. The code is ok.
Thanks anyway!

[amcqeen]

The row of labels doesn't count as an observation. If you have 1024 rows of numbers, you have 1024 entries. If you're running a VAR(2), then you can't start earlier than 3. If, in addition, you have to transform into returns within RATS, you will lose another data point doing that. If the data on the Excel file are usable as is, then you can start at 3.

Thnak you!