not significant VAR

For econometric discussions not necessarily related to EViews.

Moderators: EViews Gareth, EViews Moderator

harek
Posts: 2
Joined: Mon Jul 24, 2017 12:10 pm

not significant VAR

Postby harek » Wed Jul 26, 2017 1:09 pm

Dear all,

I am doing a VAR model on real oil price shock's effect on real GDP-growth. I have 5 different countries, and also Include the variables: Inflation, short-term rate, long-term rate, Real effective exchange rate and real wages.

As I am running my VAR, I am not able to get significant t-statistics, and my some of my coefficients are above absolute value of 1. As I have learned this is not a good fit then, and the results will be wrong.
I have tested for unit roots, and take first difference of the ones that are non stationary.
I check for optimal lag length every time, and try out several different lags.
Still, it doesn't become significant.

I have tried several different ways of transforming the data: log or log growth of real oil prices, real wages, real gdp and reer.
I have also tried to test them all in level growth rate.

As for the data, I download them from OECD and IFS.
The Oil price is deflated by the US PPI.
Real wages is deflated by individual CPI

If anyone has any ideas, please help :)

Return to “Econometric Discussions”

Who is online

Users browsing this forum: No registered users and 16 guests