Hi all,
I'm carrying out a project based on forecasting excess stock returns. 5 out of 11 of my independent variables are I(1) and my dependent variable is I(0) - determined by the ADF test.
If my dependent variable is I(0) is there point in conducting an VECM?
Thanks
Search found 5 matches
- Sun Mar 16, 2014 8:24 am
- Forum: Econometric Discussions
- Topic: Stationary dependent and VECM
- Replies: 0
- Views: 1372
- Sun Mar 16, 2014 8:22 am
- Forum: Econometric Discussions
- Topic: Basic ARCH question
- Replies: 1
- Views: 1857
Re: Basic ARCH question
From the z statistic and p-values you can see the at squared variance of the previous period is not statistically different from zero, however it is significant in determining the variance of this period. Your AIC and Schwarz criterion are quite high which suggests it won't have a good predictive po...
- Wed Feb 19, 2014 1:03 pm
- Forum: Data Manipulation
- Topic: Dating a series
- Replies: 4
- Views: 2785
Re: Dating a series
Thank you that's perfect!
- Mon Feb 17, 2014 5:51 am
- Forum: Data Manipulation
- Topic: Dating a series
- Replies: 4
- Views: 2785
Re: Dating a series
So when I open up a series it says the dates in the spreadsheet in the observation column, instead of just the number of the observation i.e. 1,2,3,4 etc.
- Sun Feb 16, 2014 12:56 pm
- Forum: Data Manipulation
- Topic: Dating a series
- Replies: 4
- Views: 2785
Dating a series
Hi, I'm trying to assign dates to my series using codes only (I can't use the interface). I've tried: %fstart "1986m2" %fstop "2014m1" smpl %first %last but every time I try that it comes back with insufficient observations in a matrix I have and all the subsequent regressions. T...
