I am trying to use EViews to replicate the article that used the normalized CES supply system from Klump, McAdam, and Willman (2007).
When I set technology progress as an exponential function, I can easily get the results of the parameters. However, after I transform the technology progress into the Box-Cox form, it gives errors such as '^ function given an out of range argument'. I am not sure how to deal with this and hope to get some help. Here is my EViews code:
@inst log_n(-1) log_y(-1) log_k(-1)
param sigma(1) 1.5 gamma_k(1) 0.01 gamma_n(1) 0.05 xi(1) 1 lambda_k(1) 1 lambda(1) 1
log(yn/xi(1)) =log((thetabar*(exp(gamma_k(1)*((tn^lambda_k(1)-1)/lambda_k(1))tbar)kn)^((sigma(1)-1)/sigma(1))+(1-thetabar)(exp(gamma_n(1)((tn^lambda_n(1)-1)/lambda_n(1))*tbar)*nn)^((sigma(1)-1)/sigma(1)))^(sigma(1)/(sigma(1)-1)))+[ar(1)=c(1)]
log_theta_n=log((1-thetabar)(exp(gamma_n(1)((tn^lambda_n(1)-1)/lambda_n(1))tbar)(xi(1)*nn/yn))^((sigma(1)-1)/sigma(1)))+[ar(1)=c(2)]
log(theta)=log(thetabar*(exp(gamma_k(1)*((tn^lambda_k(1)-1)/lambda_k(1))tbar)(xi(1)*kn/yn))^((sigma(1)-1)/sigma(1)))+[ar(1)=c(3)]
Thanks for any supportive :)
