On Mon, 17 Jan 2011, [ISO-8859-1] Alejandro Mosi�o wrote:
> Maybe i was not too much specific last time:
> I have a variable "y" that follows a GARCH(1,1) process. Then, i Gretl i
> garch 1 1 ; y const
> Then i got the result and forecasting the out-of-sample values of y can
> be done in the usual way. However, i'm interested in forecasting the
> out-of-sample variance. I don't know if such a function exists in Gretl.
There is no built-in function to do this, but you can compute a
one-step ahead forecast of the variance from the model data, as
hown in the following example script.
garch 1 1 ; Y
series e = $uhat
series h = $h
dataset addobs 10
a0 = $coeff
a1 = $coeff
b1 = $coeff
series hfc = h
# set future errors to their expectation
e = misszero(e)
# forecast the variance
hfc = a0 + a1 * e(-1)^2 + b1 * hfc(-1)
smpl 1970 ;
print e h hfc --byobs
Deal Gretl Community,
I am trying to replicate the examples in Principles of
Econometrics,pg.373, BYD example, using Garch in Gretl (GIG). I was able
to reproduce those result with Lee Adkins scripts but I cant do the same
My questions are:
- what I'm suppose to put in mean and variance regressors list to obtain
the following models:
TGARCG(1,1) and Garch in Mean.
- are there any other manual or examples using GIG gui beside those
found on Ricardo Lucchetti's home page.
Thank you for your help.
Mihai Cociuba, PhD Student, Babes-Bolyai University.
On 03/07/2011 07:00 PM, gretl-users-request(a)lists.wfu.edu wrote:
> Send Gretl-users mailing list submissions to
> To subscribe or unsubscribe via the World Wide Web, visit
> or, via email, send a message with subject or body 'help' to
> You can reach the person managing the list at
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of Gretl-users digest..."
> Today's Topics:
> 1. Odd Behavior with Saving Data, Sessions, and Scripts
> (Henrique Andrade)
> 2. Re: Help on simultaneo?us equation model (Allin Cottrell)
> 3. Help on simultaneous?us equation model (Lemma Tenessa)
> 4. Re: Odd Behavior with Saving Data, Sessions, and Scripts
> (Allin Cottrell)
> 5. Re: Help on simultaneous?us equation model (Allin Cottrell)
> Message: 1
> Date: Sun, 6 Mar 2011 23:37:53 -0300
> From: Henrique Andrade<henrique.coelho(a)gmail.com>
> Subject: [Gretl-users] Odd Behavior with Saving Data, Sessions, and
> To: "Gretl Discussion List (users)"<gretl-users(a)lists.wfu.edu>
> Content-Type: text/plain; charset="iso-8859-1"
> Dear Gretl Community,
> When I try to save data (or session, or script) files with dots in the
> names, using the GUI, Gretl assumes that all the comes after that point is
> an extension name. Suppose I have to create three files with names that help
> me to associate them with Chapter 2.5 of my book:
> Data Chapter 2.5
> Session Chapter 2.5
> Script Chapter 2.5
> I all the examples above Gretl creates the respective file with the
> extension ".5", instead of ".gdt", ".gretl", and ".inp", respectively.
> Is this the intended behavior?
> Um abra?o,
On Thu, 31 Mar 2011, Henrique Andrade wrote:
> Dear Allin, today I realized that Gretl are using the Schwert's
> suggestion to choose the maximum lag length. Thanks for
> implementing this!
> I would like to suggest to put this information into the
> Gretl-help. In the fractional integration test help we have:
> "If the optional order argument is not given the order for the
> test(s) is set automatically as the lesser of T/2 and T^0.6."
> So I think it would be nice if we could put something like that
> into the ADF's help:
> "If the lag order argument is not given the order for the test
> is set automatically as suggested by Schwert (1989):
> Max ADF lag = int(12*($nobs/100)^(1/4))"
The trouble with that is that, unlike fractint, it's not possible
to "not give" a lag order with the adf command: it's a required
argument. At present, the automatic setting is just a convenience
in the GUI. I suppose we could implement something like: give
-1 for the order if you want it set automatically a la Schwert.
On Thu, 31 Mar 2011, Mike Estrada wrote:
> I'm a new user of gretl and fairly new to econometrics in general so please
> try not to be too technical. I am using gretl to do some GARCH(1,1)
> modeling for a data verification project with my Master's Thesis. I have a
> set of Euro-Dollar exchange data that I am trying out the GARCH modeling on
> to get to know the software. My problem is that when I attempt to do this,
> I get an error that reads "Convergence criteria not met" and I have no idea
> what this means.
Gretl can't find a well-defined maximum for the likelihood within
the admissable parameter space.
One point to note is that series to be analysed via GARCH should
have a constant conditional mean -- they shouldn't be trending,
after controlling for any included regressors. With exchange
rates, which are likely to trend based on differential inflation
rates, one would either include variables that account for the
trend or perhaps run the analysis in first-differences of logs.
I'm a new user of gretl and fairly new to econometrics in general so please
try not to be too technical. I am using gretl to do some GARCH(1,1)
modeling for a data verification project with my Master's Thesis. I have a
set of Euro-Dollar exchange data that I am trying out the GARCH modeling on
to get to know the software. My problem is that when I attempt to do this,
I get an error that reads "Convergence criteria not met" and I have no idea
what this means. Could anyone please explain what this is telling me and
how I might go about fixing it? Thanks in advance.
Mike 'Chips' Estrada
On Thu, 17 Mar 2011, Henrique Andrade wrote:
> I would suggest a change in the ADF test (through the GUI). When Gretl asks
> for the lag order I think it would be nice to put a default value for the
> lag order (also leaving the option "test down from maximum lag order")
> selected by default.
> Looking at EViews (ADF test) and R (Phillips-Perron test), I'd discovered
> that they use the following default, as suggested by Schwert (Tests for Unit
> Roots: A Monte Carlo Investigation, Journal of Business & Economic
> Statistics, 1989, p.9):
> *scalar adf_lag = int(12*($nobs/100)^(1/4))*
I'm not opposed to doing something like that. What do others
think? (Right now we just offer a default of 1 for the ADF
lag-length, and leave it up to the user to raise the order, and to
select the option of testing down from some maximum order.)
I have been using Gretl recently to estimate proportion of staff, I have
three groups and the sum of the three proportion equals to 1. According to
Wooldridge (2002) this can of "share" systems can be estimated by SUR models
with Feasible GLS estimation (p. 167). Accordingly I can remove one
This is one of the models I estimate
"Skill-Mix Gap" <- system
equation pNHDfteHT2 const cvnbbeds avgnbbeds nbhospital Found TypeHT_1
TypeHT_2 TypeHT_4 TypeHT_5 gapNCo0305recod gapHCo0305recod gapMCo0305recod
equation pHDNfteHT2 const cvnbbeds avgnbbeds nbhospital Found TypeHT_1
TypeHT_2 TypeHT_4 TypeHT_5 gapNCo0305recod gapHCo0305recod gapMCo0305recod
endog pNHDfteHT2 pHDNfteHT2
instr const cvnbbeds avgnbbeds nbhospital Found TypeHT_1 TypeHT_2 TypeHT_4
TypeHT_5 gapNCo0305recod gapHCo0305recod gapMCo0305recod
estimate "Skill-Mix Gap" method=sur --iterate
The estimation is computed, no problems
I put iterate but the estimation is done with just one iteration. In the
command GRETL reference book I read that with iterate the estimation is the
maximum likelihood one.
I am completely lost because I would like to know how does that work, I
would prefer to have FGLS as in Wooldridge (my reference book in
There is not so much information around (or at least I have not found it)
about how SUR are being estimated.
If someone could help me, that would be great, I am presenting tomorrow and
I am a bit nervous,
Afer estimating ARIMA, It is convenient to test if residual is white noise in Gretl.
But I have two questions:
1. Gretl provide Ljung-Box Q statistic, residual acf and pacf.
Some paper test lag of 1 to 24,
Some only test lag of 6, 12, 18, 24.
Some only test lag of 1, 12, 24.
I am confused about the lag of Q statistic, acf and pacf.
2. Some paper test ARIMA residual by Q statistic, residual acf and pacf.
But some just test by Q statistic. Ignoring residual acf and pacf.
Are residual acf and pacf necessary when testing ARIMA residual ?
Thanks a lot