6:28 p.m.
Hi! I have a question regarding VECM and IRF interpretation in Gretl.
According to the speed of adjustment coefficient in the VEC, the target
variable adjusts at a speed of 15,5% per year to equilibrium. But the IRF
says that the cumulative change in the target variable is only 2% from year
four trough the time horizon of the IRF, which is nine years. How can I
interpret this?
Best regards,
Lars
10:14 a.m.
Am 27.04.2018 um 20:28 schrieb lasses skola:
Hi! I have a question regarding VECM and IRF interpretation in Gretl.
Hi, to be clear, this is not about an interpretation _in Gretl_, it's a
more generic question, there's nothing gretl specific here.
Which -I presume- is why no answers have followed so far.
According to the speed of adjustment coefficient in the VEC, the
target
variable adjusts at a speed of 15,5% per year to equilibrium. But the
IRF says that the cumulative change in the target variable is only 2%
from year four trough the time horizon of the IRF, which is nine years.
How can I interpret this?
This is a system there is no such thing as *the* target variable.
Apart from that, the two magnitudes have very little in common. You
could have a case where the first variable is not reacting/adjusting at
all to disequilibria, and still you could get a long-run permanent
effect because of the other dynamics.
cheers,
sven
3:47 p.m.
If I may, I would strongly recommend Professor Mallory time series blog. These and other
issues are discussed in a very clear and comprehensive manner:
http://blog.mindymallory.com/2018/02/basic-time-series-analysis-model-cho...
Tomas Nilsson
On Apr 29, 2018, at 4:14 AM, Sven Schreiber <svetosch(a)gmx.net>
wrote:
> Am 27.04.2018 um 20:28 schrieb lasses skola:
> Hi! I have a question regarding VECM and IRF interpretation in Gretl.
Hi, to be clear, this is not about an interpretation _in Gretl_, it's a more generic
question, there's nothing gretl specific here.
Which -I presume- is why no answers have followed so far.
> According to the speed of adjustment coefficient in the VEC, the target variable
adjusts at a speed of 15,5% per year to equilibrium. But the IRF says that the cumulative
change in the target variable is only 2% from year four trough the time horizon of the
IRF, which is nine years. How can I interpret this?
This is a system there is no such thing as *the* target variable.
Apart from that, the two magnitudes have very little in common. You could have a case
where the first variable is not reacting/adjusting at all to disequilibria, and still you
could get a long-run permanent effect because of the other dynamics.
cheers,
sven
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
3:43 p.m.
I'm sorry but your answer is not helping me much, it only raises new
questions for me. WHAT does the speed of adjustment say about real changes
in the variable I am interested in? WHAT is equilibrium in this context,
more than just an abstract theoretical construct? HOW do I interpret the
changes in the IRF?
More specifically, I would like to know what unit the changes in the IRF
measured in. Is it one standard deviation shock in the independent variable
on the dependent variable, measured in the unit of a standard deviation in
the dependent variable of the independent variable? Yes, I know a VAR/VEC
is a system etc.
Cheers,
Lars
2018-04-29 12:14 GMT+02:00 Sven Schreiber <svetosch(a)gmx.net>:
Am 27.04.2018 um 20:28 schrieb lasses skola:
> Hi! I have a question regarding VECM and IRF interpretation in Gretl.
>
Hi, to be clear, this is not about an interpretation _in Gretl_, it's a
more generic question, there's nothing gretl specific here.
Which -I presume- is why no answers have followed so far.
According to the speed of adjustment coefficient in the VEC, the target
> variable adjusts at a speed of 15,5% per year to equilibrium. But the IRF
> says that the cumulative change in the target variable is only 2% from year
> four trough the time horizon of the IRF, which is nine years. How can I
> interpret this?
>
This is a system there is no such thing as *the* target variable.
Apart from that, the two magnitudes have very little in common. You could
have a case where the first variable is not reacting/adjusting at all to
disequilibria, and still you could get a long-run permanent effect because
of the other dynamics.
cheers,
sven
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
7:46 p.m.
Lars,
Your questions are about VARs in general, not about Gretl. This list is for questions
specific to Gretl, not general questions about econometric methods. If you’re asking about
how Gretl computes things like IRFs, for example, that’s covered in chapter 29 of the
user’s guide. For questions about the methods, please consult other sources (e.g.,
Enders’s textbook, the blog that was recommended to you earlier, etc.).
Best,
Peter
From: gretl-users-bounces(a)lists.wfu.edu [mailto:gretl-users-bounces@lists.wfu.edu] On
Behalf Of lasses skola
Sent: Monday, April 30, 2018 11:43 AM
To: Gretl list
Subject: Re: [Gretl-users] VECM and IRF interpretation
I'm sorry but your answer is not helping me much, it only raises new questions for me.
WHAT does the speed of adjustment say about real changes in the variable I am interested
in? WHAT is equilibrium in this context, more than just an abstract theoretical construct?
HOW do I interpret the changes in the IRF?
More specifically, I would like to know what unit the changes in the IRF measured in. Is
it one standard deviation shock in the independent variable on the dependent variable,
measured in the unit of a standard deviation in the dependent variable of the independent
variable? Yes, I know a VAR/VEC is a system etc.
Cheers,
Lars
2018-04-29 12:14 GMT+02:00 Sven Schreiber
<svetosch@gmx.net<mailto:svetosch@gmx.net>>:
Am 27.04.2018 um 20:28 schrieb lasses skola:
Hi! I have a question regarding VECM and IRF interpretation in Gretl.
Hi, to be clear, this is not about an interpretation _in Gretl_, it's a more generic
question, there's nothing gretl specific here.
Which -I presume- is why no answers have followed so far.
According to the speed of adjustment coefficient in the VEC, the target variable adjusts
at a speed of 15,5% per year to equilibrium. But the IRF says that the cumulative change
in the target variable is only 2% from year four trough the time horizon of the IRF, which
is nine years. How can I interpret this?
This is a system there is no such thing as *the* target variable.
Apart from that, the two magnitudes have very little in common. You could have a case
where the first variable is not reacting/adjusting at all to disequilibria, and still you
could get a long-run permanent effect because of the other dynamics.
cheers,
sven
_______________________________________________
Gretl-users mailing list
Gretl-users@lists.wfu.edu<mailto:Gretl-users@lists.wfu.edu>
http://lists.wfu.edu/mailman/listinfo/gretl-users<https://linkprotect....
8:33 p.m.
Am 30.04.2018 um 21:46 schrieb Summers, Peter:
Your questions are about VARs in general, not about Gretl. This list
is
for questions specific to Gretl, not general questions about econometric
methods. If you’re asking about how Gretl computes things like IRFs, for
example, that’s covered in chapter 29 of the user’s guide. For questions
about the methods, please consult other sources (e.g., Enders’s
textbook, the blog that was recommended to you earlier, etc.).
I completely agree with Peter, but it has to be acknowledged that the
question about whether it's a SD shock or a unit shock is absolutely
relevant in the gretl context.
Please check out the thread "impulse-response functions from VECM in
Grtel" on this list in August 2017.
cheers,
sven
10:01 p.m.
On Mon, 30 Apr 2018, lasses skola wrote:
More specifically, I would like to know what unit the changes in the
IRF
measured in. Is it one standard deviation shock in the independent variable
on the dependent variable, measured in the unit of a standard deviation in
the dependent variable of the independent variable? Yes, I know a VAR/VEC
is a system etc.
Hmm, deja vu! You got a definitive response on just this point eight
months ago:
http://lists.wfu.edu/pipermail/gretl-users/2017-August/012630.html
Allin Cottrell
11:19 a.m.
Oh, that is true. Sorry.
2018-05-01 0:01 GMT+02:00 Allin Cottrell <cottrell(a)wfu.edu>:
On Mon, 30 Apr 2018, lasses skola wrote:
More specifically, I would like to know what unit the changes in the IRF
> measured in. Is it one standard deviation shock in the independent
> variable
> on the dependent variable, measured in the unit of a standard deviation in
> the dependent variable of the independent variable? Yes, I know a VAR/VEC
> is a system etc.
>
Hmm, deja vu! You got a definitive response on just this point eight
months ago:
http://lists.wfu.edu/pipermail/gretl-users/2017-August/012630.html
Allin Cottrell
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
2179
days inactive
2184
days old
7 comments
5 participants
participants (5)
-
Allin Cottrell -
lasses skola -
Summers, Peter -
Sven Schreiber -
Tomas Nilsson