cointegration bound test
Posted: Tue Dec 31, 2013 3:42 am
i have 8 variables , which are : lnrgd cg_y cp_y f_y hg_y ip_y ig_y x_y . four of them are I(0) and the remaining variables are I(1) , so i want to do bound test for checking the long- run relationship between these variables . i do some steps but i do not know if they are correct or not ,in addition i do not know what should i do after that ?.
first i estimate this equation :
d(lnrgdp) lnrgdp(-1) cg_y(-1) cp_y(-1) f_y(-1) hg_y(-1) ig_y(-1) ip_y(-1) x_y(-1) d(lnrgdp(-1)) d(cg_y(-1)) d(cp_y(-1)) d(f_y(-1)) d(hg_y(-1)) d(ig_y(-1)) d(ip_y(-1)) d(x_y(-1)) c
i want to know if this eqation is right or wrong .
then i do wald test which null hypothesis is C(1)= C(2)=C(3)=C(4)=C(5)=C(6)=C(7)=C(8)=0
the p- value of f was 0.3531 ( which means that there is no cointegration ).
i want to know if these steps are true or false .
when i do diagnostic tests , before or after wald test ? especially i notice that the all t coefficients are insignificant ? and what are the steps after that ?
i need a help , i am a biginner with cointergration , so i wish if somebody explain the steps in details for doing this test .
note : i use eviews 7
first i estimate this equation :
d(lnrgdp) lnrgdp(-1) cg_y(-1) cp_y(-1) f_y(-1) hg_y(-1) ig_y(-1) ip_y(-1) x_y(-1) d(lnrgdp(-1)) d(cg_y(-1)) d(cp_y(-1)) d(f_y(-1)) d(hg_y(-1)) d(ig_y(-1)) d(ip_y(-1)) d(x_y(-1)) c
i want to know if this eqation is right or wrong .
then i do wald test which null hypothesis is C(1)= C(2)=C(3)=C(4)=C(5)=C(6)=C(7)=C(8)=0
the p- value of f was 0.3531 ( which means that there is no cointegration ).
i want to know if these steps are true or false .
when i do diagnostic tests , before or after wald test ? especially i notice that the all t coefficients are insignificant ? and what are the steps after that ?
i need a help , i am a biginner with cointergration , so i wish if somebody explain the steps in details for doing this test .
note : i use eviews 7