[no subject]

["Leyla Khoury" <lillie_pancake@hotmail.com>]

Hi all,

 

Sorry for the previous post, had trouble with my account.

 

I am writing a program in S+GARCH which fits an MGARCH model (with the BEKK specification for the conditional second moment) to a specified data set. I use the command

garch_mgarch(e1e2 ~ arma(0,0), ~ bekk(1,1))

I then would like to fit an exogenous variable (the Error Correction Term squared) and use the command

    garchx_mgarch(e1e2 ~ arma(0,0), ~ bekk(1,1) + ectsq)

The estimates for each of the four additional parameters are 0, yet the estimates of the other 13 parameters do not equal the parameter estimates when only the GARCH model is fitted (I have defined ectsq as a matrix). I am unsure if there is an error in the code or the importation of the exogenous data set.

Thank you for any help offered,

Liliana

 

 

 

 

---

Send and receive Hotmail on your mobile device: Click Here