s-news
[Top] [All Lists]

Fw: FIEGARCH

To: <s-news@lists.biostat.wustl.edu>
Subject: Fw: FIEGARCH
From: "Bikash Jain" <bjain@bslindia.com>
Date: Thu, 23 Jun 2005 09:48:46 +0530
 

Hi All,

I am attaching the data set. The data set contains daily return series (called ?Variable?). There are more than 3900 data points in the series.

Let me again mention that, we could run GARCH, EGARCH and FIEGARCH models on this data set. But, when we give the FIGARCH command [Pjohn.figarch= fgarch(data.file.Pjohn[["Variable"]]~1+ma(1),~figarch(1,1))], we get the following message

?Initializing model parameters.

Iteration No.   1: log-likelihood=10805.156171

Routine aborted. Try rescaling the series or a different initialization.?

 

However, when we use FIEGARCH command[Pjohn.fiegarch= fgarch(data.file.Pjohn[["Variable"]]~1+ma(1),~fiegarch(1,1), leverage=T)], we get the following output:

?Initializing model parameters.

Iteration No.   1: log-likelihood=10784.179192

Iteration No.   2: log-likelihood=10799.276870

???????

Iteration No.  18: log-likelihood=10809.027524

Iteration No.  19: log-likelihood=10809.027524

Convergence in function value.?

 

Thus, the FIEGARCH model could converge, whereas, FIGARCH iteration got aborted after first iteration. We could not find out the reason for this. Can you help?  

Alternatively, are you aware of any statistical software that can model FIGARCH and SV ? We urgently require this information so that we can finish our research.

 

Waiting for your quick response.

 

Bikash Jain

<Prev in Thread] Current Thread [Next in Thread>
  • Fw: FIEGARCH, Bikash Jain <=