EViews.com

Hello,

For my master thesis i am currently using EViews 7 SV to do my research. I have been busy with it quite some time now, but unfortunatly i am having some errors occuring and to be honest i am having trouble solving them and getting a bit frustrated by it.
As can be seen in my attached workfile, i have a number of variables, all of which i took first difference or logs etc. Now my research take a look at the logs and i therefore created the groups "all_lg_us" and "all_lg_ger". There are various tests i need to do, all which are going fine in the group "all_lg_us". However, when going to the "all_lg_ger" group i am having trouble already at the first test i am performing, the Johansen Cointegration test. It keeps telling me that there is a Near Singular Matrix. I have been trying different things for two days now, none of which seem to solve the problem. Any suggestions on what can be done to solve? Is there a solution?
The second problem that i am having is that when trying to create a VECM of this group, it keeps telling me i havent got enough observation, this seems odd to me since i do actually got quite some few observations, how can this be solved?
The frustrating part of it all is that everything went perfectly fine with the first group "all_lg_us" and the moment i start on the other group it keeps giving me error signals.

Anyone that could help me? It would be highly appreciated since it is getting the better of me :?

Kind regards

You only have 30 observations for the Germany group. That isn't enough to do the things you want to do.

with the german group, do you mean i need more data for my variables trade_balance_ger and gdp_germany? I think only the variable trade_balance_ger has not enough observations then, am i correct?

Does more german data solve the near singular problem too??