[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: simultaneously GARCH

From	"Jonathan" <[email protected]>
To	<[email protected]>
Subject	st: simultaneously GARCH
Date	Fri, 5 Aug 2005 14:09:01 +0100

Statlist,
 
 
I am trying to deal with the following GARCH ;
The general form of the Multivariate GARCH(p,q)-M is the following:
cid:[email protected]
with
cid:[email protected]
where y
t 
is the Nx1 vector of excess returns of all the assets in the market and
           H
t
is the covariance matrix of the excess returns given information available at
                time t-1.
We want to estimate the BIVARIATE GARCH(1,1)-M model of the form:
cid:[email protected]
 where R
i
and R
m
are the elements of the y
t
vector
cid:[email protected]
 
cid:[email protected]
    where
cid:[email protected]
 is the variance of u
t
cid:[email protected]
  where
cid:[email protected]
 is the variance of v
t
cid:[email protected]
 (assumption of constant conditional correlation)
 
We need to find the estimates of
&agr;
,
&lgr;
,
&bgr;
,
&ggr;
,
&ggr;
1
,
&ggr;
2
,
&dgr;
,
&dgr;
1
,
&dgr;
2
,
&rgr;
 and to compute the time varying beta coefficient
&bgr;
i
= Cov
t
(R
i
,R
m
)/Var
t
(R
m
).
 
Is there any way STATA let you estimate a
multivariate or even a Bivariate GARCH-M simultaneously? (you can estimate the
equations, but only one at a time and not simultaneously)
 
best,
 
jon