Page 1 of 1

lags of variance

Posted: Mon Mar 04, 2013 4:56 am
by strypste
Hi

Is there an way to include lags of the variance in the mean equation in a ARCH-M model?

So, for example, the model could be:

y_t = c + lambda1*sigma_t + lambda2*sigma_t-1 + e_t

where sigma^2_t = w0 + w1*e^2_t-1.

For now, I use the command: equation ARCH1M.arch(1,0,archm=sd) c

This estimates an ARCH1 model with sigma_t in the mean equation and with ARCH(1) errors.


Thanks...