R help - Nov 2013 - Test for exogeneity

Hi,



   I am building a bivariate SVAR model



y_1t=c_1+Ã_1 (1,1) y_(1,t-1)+Ã_1 (1,2) y_(2,t-1)+Ã_2 (1,1) y_(1,t-2)+Ã_2
(1,2) y_(2,t-2)+å_1t



   b y_1t+ y_2t=c_2+Ã_1 (2,1) y_(1,t-1)+Ã_1 (2,2) y_(2,t-1)+Ã_2 (2,1)
y_(1,t-2)+Ã_2 (1,2) y_(2,t-2)+å_2t



  Now y1 is "relatively" exogenous in that y1 impacts y2
contemporaneously
but not the other way around. Given a bivariate dataset, is there any
statistical test (in any R package or elsewhere) that helps to justify/test
the exogeneity of y1 in the present context? Is there any reference
available?



Thanks,



Miao

	[[alternative HTML version deleted]]