Granger Causality Test - Help needed

[zobee]

Hi, i'm doing my thesis for MSc and am newbie regarding eview. I need help regarding granger causality test. basically my research involve two main variables - real estate and stock exchange, with three exogenous variables. the problem is that with the built-in granger test function in eviews, there is no option for including exogenous variables, it only asks for no. of lags. i want to use both linear and non-linear granger test. so, i need help regarding how to carry this test. i presume i'll have to use VAR equation function (correct me if i'm wrong), but i dont understand how to write the equation (or matter of fact where to write it in eviews). also, i'll not be using error correction model for VAR as no significant cointegration is found between both variables, so i need to do only granger test. specifically, i need help regarding how to write the equation say if my main two variables are RE and SE (where RE is independent and SE is dependent variable), with three exogenous variables x1, x2, and x3? can anyone help me, thanks in advance.

[theologos]

Hello,

In your E-views work-file, select the RE and SE variables. Open them as a VAR and at bottom (left side) in the VAR specification window, there is a field called: Exogenous Variables, where you should add the X1, X2 and X3 (The constant is by default present). After the estimation, to perform the Granger-causality test, simply test the joint significance of the lags.

Regards

[zobee]

thanks, its really helpful. do you know something regarding linear & non-linear granger causailty test? i mean how to differently run both tests?

[theologos]

Hi,

The above Granger causality test is linear. E-views do not perform a non-linear Causality tests. I am afraid that you need to construct your own code.

Regards

[zobee]

thanks, can you kindly help me in the coding because i'm trying to do it with help of eviews guide but i'm coming up with errors. also, my exogenous variables are now reduced to one variable of inflation rate (IR).

p.s. i wante to ask regarding your past post about joint significance of lags, you mean it my going to lag structure --> lag exclusion test and looking at the last cloumn which says 'joint'??? precisely what am i suppose to look from these figures? the highest or lowest one? also, there is one option of exogeniety (granger causality) test in the menu after you estimate VAR. am i suppose to look perform this test to obtain information regarding granger cause relationship? for these p-values am i suppose to look for values greater than or les than .05 to get positive answer for causality i.e. if value is greater han there exists causality or if lesser than .05? thanks in advance

[zobee]

or matter of fact, anyone else can also help me???