VAR Unit root testing
Posted: Sat Dec 01, 2012 7:43 am
Dear All,
I am new to the forum and I am currently enrolled in a empirical methods course.
I have an upcoming exam and struggle to understand how to solve a problem set provided by the lecturer.
The guy requires us to compute the unit roots of VAR model using point estimates ourselves.
I have two variables and their point estimates. Using eigen values I have to compute and test for unit roots without using Eviews ofcourse.
Could someone please show me how to do that and write down which formulas I have to plug the numbers in...
It's just a very tough course and I fail to understand the maths behind this question.
I initally thought that I could compute them simply by dividing 1 by the point estimate and if I get 1 then there is unit root. I got told by the lecture that if the value is for example 2.5 this will not work. Currently I have PAt = 0.601; PBt = 0.327; PAt-1=0.003; PBt-1 = 0.501.
I would be grateful if one could help.
Regards
I am new to the forum and I am currently enrolled in a empirical methods course.
I have an upcoming exam and struggle to understand how to solve a problem set provided by the lecturer.
The guy requires us to compute the unit roots of VAR model using point estimates ourselves.
I have two variables and their point estimates. Using eigen values I have to compute and test for unit roots without using Eviews ofcourse.
Could someone please show me how to do that and write down which formulas I have to plug the numbers in...
It's just a very tough course and I fail to understand the maths behind this question.
I initally thought that I could compute them simply by dividing 1 by the point estimate and if I get 1 then there is unit root. I got told by the lecture that if the value is for example 2.5 this will not work. Currently I have PAt = 0.601; PBt = 0.327; PAt-1=0.003; PBt-1 = 0.501.
I would be grateful if one could help.
Regards