EViews.com

Posted: **Fri Sep 15, 2017 6:19 am**

i was trying to use LOGl to estimate an EGARCH model, here is my code

@logl logl1
res = er - c(1) - c(2)*var - c(3)*var*D - c(4)*tr
LOG(var) = C(5) + C(6)*ABS(RESID(-1)/@SQRT(var(-1))) + C(7)*RESID(-1)/@SQRT(var(-1)) + C(8)*LOG(var(-1)) + C(9)*op+ C(10)*pe + C(11)*RF
logl1 = log(@dnorm(res/@sqrt(var))) - log(var)/2

However, i get this information

Missing values in @LOGL series at current coefficients at observation 6/02/1989

So, can some experts tell me why and how to solve it?

Posted: **Fri Sep 15, 2017 6:23 am**

sorry, the code is the below

Code: Select all

@logl logl1
res = er - c(1) - c(2)*var  - c(4)*tr
LOG(var) = C(5) + C(6)*ABS(RESID(-1)/@SQRT(var(-1))) + C(7)*RESID(-1)/@SQRT(var(-1)) + C(8)*LOG(var(-1)) + C(9)*op+ C(10)*pe + C(11)*RF
logl1 = log(@dnorm(res/@sqrt(var))) - log(var)/2

Posted: **Fri Sep 15, 2017 6:27 am**

here is the workfile

Posted: **Fri Sep 15, 2017 7:39 am**

the starting value is set based on

Code: Select all

equation basic.arch(ged, egarch, archm=var) er c tr @ op pe rf

Posted: **Fri Sep 15, 2017 8:20 am**

RESID has no data. Are you sure you want it to depend on RESID?

Posted: **Sat Sep 16, 2017 6:48 am**

RESID has no data. Are you sure you want it to depend on RESID?

Hi Garath, thanks for your reply.

yes. as for the GARCH (1,1), i would like to add it to the volatility model. i know that Eviews has package to do it in a simple way; however, i need to modify the mean equation in the following procedure.

so, how can i solve it? i have pre-estimated a model and i believe after that estimation, resid should have values.

Posted: **Sat Sep 16, 2017 6:57 am**

RESID has no data. Are you sure you want it to depend on RESID?
Hi Garath, thanks for your reply.

yes. as for the GARCH (1,1), i would like to add it to the volatility model. i know that Eviews has package to do it in a simple way; however, i need to modify the mean equation in the following procedure.

so, how can i solve it? i have pre-estimated a model and i believe after that estimation, resid should have values.

I mean EGARCH(1,1)

EViews.com

Likelihood function problem

Likelihood function problem

Re: Likelihood function problem

Re: Likelihood function problem

Re: Likelihood function problem

Re: Likelihood function problem

Re: Likelihood function problem

Re: Likelihood function problem