Return-Path: <maiser@efs.mq.edu.au>
Received: from corot.bc.edu (corot.bc.edu [136.167.2.209])
	by monet.bc.edu (8.8.7/8.8.7) with ESMTP id MAA105364
	for <baum@monet.bc.edu>; Fri, 2 Apr 1999 12:17:24 -0500
From: maiser@efs.mq.edu.au
Received: (from root@localhost)
	by corot.bc.edu (8.8.7/8.8.7) with X.500 id MAA72844
	for baum@mail1.bc.edu; Fri, 2 Apr 1999 12:17:24 -0500
Received: from baldrick.ocs.mq.edu.au (baldrick.ocs.mq.edu.au [137.111.1.12])
	by corot.bc.edu (8.8.7/8.8.7) with ESMTP id MAA75372
	for <baum@bc.edu>; Fri, 2 Apr 1999 12:17:19 -0500
Received: from efs1.efs.mq.edu.au (efs1.efs.mq.edu.au [137.111.64.8])
	by baldrick.ocs.mq.edu.au (8.9.2/8.9.2) with ESMTP id DAA00532
	for <baum@bc.edu>; Sat, 3 Apr 1999 03:17:09 +1000 (EST)
Received: from EFS1/SpoolDir by efs1.efs.mq.edu.au (Mercury 1.40);
    3 Apr 99 03:18:56 GMT+1000
Received: from SpoolDir by EFS1 (Mercury 1.40); 3 Apr 99 03:18:55 GMT+1000
To: baum@bc.edu
Date: Sat, 3 Apr 99 3:18:55 GMT+1000
Subject: Re: 
Message-ID: <8253D567AC5@efs1.efs.mq.edu.au>

From: jankowski@ait.fredonia.edu (Dick Jankowski)
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Unit Root Tests
Date: Sun, 28 Feb 1999 13:49:33 -0500
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
X-Mailer: Windows Eudora Version 1.4.4 (via Mercury MTS (Bindery) v1.40)
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"

Rats aficionados,
        I am using the program "Unitroot.src" to perform ADF and
Perron-Phillips unit root tests.  One of the test statistics generates two
different statistics for rho, T(rho-1) and t(rho-1).  What is the difference
between the two, or is there a source I can consult?
        Dick Jankowski
        Political Science Department
        SUNY at Fredonia


---------- End of message ----------

From: "Dr. Sean Michael Snaith" <ssnaith@uwimona.edu.jm>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Re: Unit Root Tests
Date: Mon, 01 Mar 1999 12:16:57 -0500
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
X-Mailer: QUALCOMM Windows Eudora Pro Version 4.0 (via Mercury MTS (Bindery) v1.40)
Mime-Version: 1.0
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

Dick,

I have not used the source program you are referring to, but I suspect that
the
different statistics reported in this program correspond to whether or not a
drift term and trend(centered) are included in the regression which is
estimated to give you the coefficient on the AR(1) term.  Generally the=
 three
models are: no drift or trend, drift but no trend and both drift and trend.

One source for discussion about parametric and non-parametric unit roots=
 test
is "Cointegration, Error-Correction, and the Econometric Analysis of
non-stationary Data, by Banerjee et al, Oxford University Press.


Regards,

Sean





Dr. Sean Michael Snaith
Department of Economics
University of the West Indies
Kingston, 7
Jamaica
West Indies

TEL: (876) 977-1188=A0=20
FAX: (876) 977-1483


---------- End of message ----------

From: Irida Achthoven <achthovi@qed.econ.queensu.ca>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: sample LGARCH PRG.
Date: Mon, 01 Mar 1999 15:06:29 -0800
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
Organization: Queen's University
X-Mailer: Mozilla 3.04 (Win16; I) (via Mercury MTS (Bindery) v1.40)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Hallo everybody.

Ilike to request a sample program for the lgarch model.
I have tried to to use the  latest version of the garch.src , but the
procedure seems to have a problem.
I have also tried to use the previous version, but tha in mean option
does not seem to work either.
So I hope somebody can help me
Thank you
Irida

---------- End of message ----------

From: Irida Achthoven <achthovi@qed.econ.queensu.ca>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Sample forecasting.prg
Date: Mon, 01 Mar 1999 15:10:56 -0800
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
Organization: Queen's University
X-Mailer: Mozilla 3.04 (Win16; I) (via Mercury MTS (Bindery) v1.40)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

hallo everybody,
I would like to request an sample (G)-archm forecasting prg (rolling
forecast).
It can be either a bootstrap or a Monte carlo simulation.
thank you
Irida

---------- End of message ----------

From: Norman Morin <m1njm00@frb.gov>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: GARCH
Date: Mon, 01 Mar 1999 14:44:05 -0500
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
MIME-version: 1.0
X-Mailer: exmh version 2.0.2.4 7/8/98 (via Mercury MTS (Bindery) v1.40)
Content-type: text/plain; charset=us-ascii

Um, remember when I suggested adding a few lines of
LOCAL statements to the GARCH proc.  Well, that makes it blow
up, so subtract them if you do not want it to blow up.

They were the LOCAL REAL lines around 510 lines into the proc.

I guess I ought to try a fix before suggesting it.

Norm
-- 

                           Norman J. Morin 
                  nmorin@frb.gov or m1njm00@frb.gov         
          --------------------------------------------------
          Division of Research and Statistics * Mail Stop 82
           Board of Governors of the Federal Reserve System 
               Washington, D.C. 20551 * (202) 452-2476
          --------------------------------------------------



---------- End of message ----------

From: "Guillaume, Dominique" <Dominique.Guillaume@econ.kuleuven.ac.be>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: question garch procedure
Date: Wed, 3 Mar 1999 17:15:26 +0100 
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2448.0) (via Mercury MTS (Bindery) v1.40)
Content-Type: text/plain;

Hello,

I am new in using rats and tried to run the GARCH.src procedure available on
the estima site but got
An error message mentioning I should "allocate" memory to series. Anyone who
had that - I guess- 
Fairly easy problem before with the same procedure? 

Thanks  a lot,

Dominique

---------- End of message ----------

From: khl2288@UTARLG.UTA.EDU
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: t test
Date: Wed, 03 Mar 1999 18:22:36 -0600 (CST)
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
MIME-version: 1.0
Content-type: TEXT/PLAIN; charset=US-ASCII
X-Mailer: Mercury MTS (Bindery) v1.40

Hi, RATS users.
I'd like to know how to conduct t test for the followng setting.
I have two variables, A and B, and calculate cross corrlation of the two
variables by using CROSS;
Cross A B 1 100 -6 6 crosscor

Crosscor is the saved correlation coefficients from the Cross statement.
I'd like to get the significance levels of the correlation coefficients.
I know that we can get Q-stats using qstats option, but what I want to
know is whether the correlation coefficient at each lag and lead is
significant or not. The null hypothesis would be the rho(k)=0.
I'm using RATS version 4.3.

Many thanks in advance.

Kyong H. Lee
Dept. of Economics
University of Texas at Arlington


---------- End of message ----------

From: "Kahlil Rowter" <krowter@zdnetmail.com>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Unsubscribe
Date: Thu, 04 Mar 1999 05:00:16 -0000
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
Mime-Version: 1.0
X-Mailer: MailCity Service (via Mercury MTS (Bindery) v1.40)
Organization: ZDNet Mail  (http://www.zdnetmail.com:80)
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Unsubscribe 


Free web-based email, anytime, anywhere! 
ZDNet Mail - http://www.zdnetmail.com

---------- End of message ----------

From: Manfred.Kremer@bundesbank.de
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Reply: t test
Date: 04 Mar 1999 11:28:18 +0000
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
X-Mailer: Mercury MTS (Bindery) v1.40


Dear Kyong,

I ran across the following test.

The variance of the correlation coefficient can be estimated by 1 divided
by the number of observations (nobs).
Thus, testing the significance of a correlation coefficient can simply be
based on the following t-statistic:

                            rho(k)
t-statistic =  --------------------
                       (1/nobs)**0.5

Manfred



---------- End of message ----------

From: khl2288@UTARLG.UTA.EDU
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Re: Reply: t test
Date: Sat, 06 Mar 1999 16:46:49 -0600 (CST)
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
MIME-version: 1.0
Content-type: TEXT/PLAIN; charset=US-ASCII
X-Mailer: Mercury MTS (Bindery) v1.40

Hi, Manfred.
Thanks for your help.
I have one more question. I also need to run a t test for the SQUARED two
asset returns for the significance of their cross correlations in lags=10
and leads=10. Can I use the same formula for the t test statitics you gave
me for the squared series?

Thank you so much.

Kyong H. Lee
Dept. of Economics
University of Texas at Arlington

On Thu, 4 Mar 1999 Manfred.Kremer@bundesbank.de wrote:

> 
> Dear Kyong,
> 
> I ran across the following test.
> 
> The variance of the correlation coefficient can be estimated by 1 divided
> by the number of observations (nobs).
> Thus, testing the significance of a correlation coefficient can simply be
> based on the following t-statistic:
> 
>                             rho(k)
> t-statistic =  --------------------
>                        (1/nobs)**0.5
> 
> Manfred
> 
> 
> 


---------- End of message ----------

From: KHL2288@UTARLG.UTA.EDU
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Covariance stationarity
Date: Sat, 06 Mar 1999 21:39:41 -0600 (CST)
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
MIME-version: 1.0
Content-type: TEXT/PLAIN; CHARSET=US-ASCII
X-Mailer: Mercury MTS (Bindery) v1.40

Dear RATS users,
I'm estimating BEKK specification of GARCH(1,1) process. Does anyone have a codeto test the covariance stationarity of the GARCH(1,1) process.

Thank you so much.

Kyong H. Lee
Dept. of Economics
University of Texas at Arlington

---------- End of message ----------

From: khl2288@UTARLG.UTA.EDU
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: LR test
Date: Sat, 06 Mar 1999 23:05:05 -0600 (CST)
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
MIME-version: 1.0
Content-type: TEXT/PLAIN; charset=US-ASCII
X-Mailer: Mercury MTS (Bindery) v1.40

Dear RATS users,
I have a question in conducting Likelihood Test in GARCH(1,1)-M
setting.For example, if I set a bivariate GARCH(1,1)-M model to test
whether the covariance risk in CAPM is priced;
y1=a11+b11*h11
y2=a21+b21*(h11/h12),

the null hyphothesis would be b21=0.
So, I first estimated the full model which inculdes the parameter b21, and
saved the residuals as r1 for the y1 and r2 for the y2. And then, I used 
NONLIN(drop) b21 in the estimation procedure to estimate the restricted
model and saved the residuals as r3 for y1 and r4 for the y2.
Finally, I did the following to do a LR test for the null of b21=0;
RATIO(degrees=1)
# r1 r2
# r3 r4

Am I following the right procedure?

I'll really appreciate for any comment.



Kyong H. Lee
Dept. of Economics
University of Texas at Arlington


---------- End of message ----------

From: "Krassimir Petrov" <krassimir_petrov@hotmail.com>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: RATS threads
Date: Sun, 07 Mar 1999 00:44:35 PST
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
Mime-Version: 1.0
Content-type: text/plain
X-Mailer: Mercury MTS (Bindery) v1.40

Hello everybody,

I recently tried to speed up computations and switched to a dual-pentium 
computer only to find out that there is no speed gain. The NT used only 
50% of the processor time. I checked on the RATS process and found that 
THERE IS ONLY 1 THREAD running, so using dual- or quadprocessor computer 
would not speed up things!

I ran this on exactly one partucular program based on Morin's var.src 
procedure.

My question is whether RATS is inherently single-threaded program or it 
is just the particular code that I used.


Regards to all.

Chris.


______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com

---------- End of message ----------

From: "Ziya Korkut Esrefoglu" <kesrefog@anadolu.edu.tr>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: UNSUBSCRIBE
Date: Mon, 8 Mar 1999 08:02:00 +0200
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
MIME-Version: 1.0
Content-Type: text/plain;
Content-Transfer-Encoding: 7bit
X-Mailer: Microsoft Outlook Express 5.00.0810.800 (via Mercury MTS (Bindery) v1.40)

Unsubscribe


---------- End of message ----------

From: Manfred.Kremer@bundesbank.de
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Antwort: Re: Reply: t test
Date: 08 Mar 1999 15:56:59 +0000
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
X-Mailer: Mercury MTS (Bindery) v1.40


Hi Kyong,

If you look for a reference concerning standard errors of correlation
coefficients between two (normally distributed) random variables, you
should at best consult Alan Stuart and J. Keith Orr: Kendall's Advanced
Theory of Statistics, Vol. 1, Distribution Theory, 6th edition, London et
al. 1994, pp. 355 f. and 558-564.

On p. 356 you find the formula for the variance of the correlation
coefficient r in the sample:

var(r) = (1/nobs)(1-rho**2)**2

with rho the correlation coefficient in the population.

Hence, the variance can be estimated by (1/nobs) as stated in my first mail
only under the null hypothesis rho = 0.

Since this variance estimate is only valid for normally distributed
variables, it doesn't work for your squared asset returns series. Perhaps
you will find a solution to your problem in the volume cited above, too.

Good luck!

Manfred



---------- End of message ----------

From: Irida Achthoven <achthovi@qed.econ.queensu.ca>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: MODIFY THE GARCHMV.PRG TO ALLOW FOR ARCH-IN-MEAN
Date: Tue, 09 Mar 1999 03:55:34 -0800
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
Organization: Queen's University
X-Mailer: Mozilla 3.04 (Win16; I) (via Mercury MTS (Bindery) v1.40)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

I was wondering if you could answer the following question for me,


Dear Rats user's
I need to modify the garchmv.prg to allow for arch term (conditional
variance) in the conditional mean.
I have tried to modify the model, but was not very succesful.


The problem I ran into is that  the parameter estimates for the variance
term of the mean eqn have zero sstandard deviation and therefore a zero
t-stat.
Therefore  I am sure that I am doing something wrong.
Could somebody  help me please.
I would really appreciate it.
Thank you
Irida achthoven

---------- End of message ----------

From: Hazem Daouk <hdaouk@indiana.edu>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Generalized Impulse Responses
Date: Thu, 11 Mar 1999 01:39:18 -0500 (EST)
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
X-Mailer: Mercury MTS (Bindery) v1.40

Dear RATS users,

I would like to implement the Generalized Impulse Response analysis in the
context of a VAR model. "The generalized impulse responses are new and,
unlike the orthogonalized responses, do not depend on the ordering of the
variables in unrestricted VAR models." This method was proposed in the
following two papers:

Koop, G., Pesaran, H., Potter, S. (1996) Impulse Response Analysis in
Nonlinear Multivariate Models, Journal of Econometrics, vol 74, p 119-147.

Pesaran, H., Shin, Y. (1998) Generalized Impulse Response Analysis in
Linear Multivariate Models, Economics Letters, vol 58, p 17-29.

I know that the software MicroFit can compute and plot generalized impulse
responses. However, I would like to stick with RATS since that's where all
my programs are. I was wondering if somebody has a RATS code that
implements generalized impulse responses or can offer some advice about
this issue? 

Thanks in advance,

Hazem Daouk
Finance Department
Kelley School of Business
Indiana University
Bloomington, IN 47406




---------- End of message ----------

From: "Manfred Stamer" <STAMER2@hermes1.econ.uni-hamburg.de>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: unsubscribe
Date: Thu, 11 Mar 1999 17:23:17 GMT+0100
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
X-mailer:     Pegasus Mail/Windows (v1.11a) (via Mercury MTS (Bindery) v1.40)

Sorry,
I don't remember the right address for unsubscribing.
So I choose the way.

Best regards,
Manfred
 
Manfred Stamer
Hohensteiner Strasse 7
D-22049 Hamburg

e-mail stamer2@hermes1.econ.uni-hamburg.de

---------- End of message ----------

From: KHL2288@UTARLG.UTA.EDU
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Re: MODIFY THE GARCHMV.PRG TO ALLOW FOR ARCH-IN-MEAN
Date: Thu, 11 Mar 1999 10:46:12 -0600 (CST)
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
MIME-version: 1.0
Content-type: TEXT/PLAIN; CHARSET=US-ASCII
X-Mailer: Mercury MTS (Bindery) v1.40

Hi, Irida.
Well, the following is a part of my program for ARCH-in-M. But, I am not sure this right one even though everything looks fine to me.

Nonlin b11 b12 b21 b22
Frml res1 = y1 - b11 - b12*y1{1}
Frml res2 = y2 - b21 - b22*(h12/h11)

It may not be the form of ARCH-in-Mean you want. This is a bivariate model for,say,CAPM. Everything else in GARCHMV.PRG is unchanged.

I wish it helps you a little.

Kyong H. Lee
Dept. of Economics
University of Texas at Arlington

---------- End of message ----------

From: "Iacoviello,M  (pg)" <M.Iacoviello@lse.ac.uk>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: RE: Generalized Impulse Responses
Date: Thu, 11 Mar 1999 16:53:40 -0000
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2448.0) (via Mercury MTS (Bindery) v1.40)
Content-Type: text/plain

You'd better use Microfit, which is the program written by Pesaran and does
all these things.
It is also much simpler than RATS
Matteo
> ----------
> From: 	Hazem Daouk[SMTP:hdaouk@indiana.edu]
> Reply To: 	RATS Discussion List
> Sent: 	11 March 1999 06:39
> To: 	RATS Discussion List
> Subject: 	Generalized Impulse Responses
> 
> Dear RATS users,
> 
> I would like to implement the Generalized Impulse Response analysis in the
> context of a VAR model. "The generalized impulse responses are new and,
> unlike the orthogonalized responses, do not depend on the ordering of the
> variables in unrestricted VAR models." This method was proposed in the
> following two papers:
> 
> Koop, G., Pesaran, H., Potter, S. (1996) Impulse Response Analysis in
> Nonlinear Multivariate Models, Journal of Econometrics, vol 74, p 119-147.
> 
> Pesaran, H., Shin, Y. (1998) Generalized Impulse Response Analysis in
> Linear Multivariate Models, Economics Letters, vol 58, p 17-29.
> 
> I know that the software MicroFit can compute and plot generalized impulse
> responses. However, I would like to stick with RATS since that's where all
> my programs are. I was wondering if somebody has a RATS code that
> implements generalized impulse responses or can offer some advice about
> this issue? 
> 
> Thanks in advance,
> 
> Hazem Daouk
> Finance Department
> Kelley School of Business
> Indiana University
> Bloomington, IN 47406
> 
> 
> 

---------- End of message ----------

From: "Masih, Rumi" <Rumi.Masih@gs.com>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: RE: Generalized Impulse Responses
Date: Thu, 11 Mar 1999 12:28:03 -0500
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.0.1460.8) (via Mercury MTS (Bindery) v1.40)
Content-Type: text/plain

I strongly support this idea. There are some procedures run in Gauss which
just about can replicate anything in menu driven packages but your best bet
is Microfit. I am sure its possible to generate GIRFs in RATs but of you are
not interested in computing any standard error bands Microfit would be the
easiest route.

_____________________________________
Rumi Masih
Emerging Markets Economic Research
Goldman, Sachs & Co.
85 Broad Street, New York  NY  10004
USA
Ph: 1-212-902-4410
Fax: 1-212-346-4402

> -----Original Message-----
> From:	Iacoviello,M  (pg) [SMTP:M.Iacoviello@lse.ac.uk]
> Sent:	Thursday, March 11, 1999 11:54 AM
> To:	RATS Discussion List
> Subject:	RE: Generalized Impulse Responses
> 
> You'd better use Microfit, which is the program written by Pesaran and
> does
> all these things.
> It is also much simpler than RATS
> Matteo
> > ----------
> > From: 	Hazem Daouk[SMTP:hdaouk@indiana.edu]
> > Reply To: 	RATS Discussion List
> > Sent: 	11 March 1999 06:39
> > To: 	RATS Discussion List
> > Subject: 	Generalized Impulse Responses
> > 
> > Dear RATS users,
> > 
> > I would like to implement the Generalized Impulse Response analysis in
> the
> > context of a VAR model. "The generalized impulse responses are new and,
> > unlike the orthogonalized responses, do not depend on the ordering of
> the
> > variables in unrestricted VAR models." This method was proposed in the
> > following two papers:
> > 
> > Koop, G., Pesaran, H., Potter, S. (1996) Impulse Response Analysis in
> > Nonlinear Multivariate Models, Journal of Econometrics, vol 74, p
> 119-147.
> > 
> > Pesaran, H., Shin, Y. (1998) Generalized Impulse Response Analysis in
> > Linear Multivariate Models, Economics Letters, vol 58, p 17-29.
> > 
> > I know that the software MicroFit can compute and plot generalized
> impulse
> > responses. However, I would like to stick with RATS since that's where
> all
> > my programs are. I was wondering if somebody has a RATS code that
> > implements generalized impulse responses or can offer some advice about
> > this issue? 
> > 
> > Thanks in advance,
> > 
> > Hazem Daouk
> > Finance Department
> > Kelley School of Business
> > Indiana University
> > Bloomington, IN 47406
> > 
> > 
> > 

---------- End of message ----------

From: Galindo Andrade Arturo <agalinan@banrep.gov.co>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: RATIO of MA/AR Polinomials
Date: Fri, 12 Mar 1999 09:48:23 -0500
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2448.0) (via Mercury MTS (Bindery) v1.40)
Content-Type: text/plain;
Content-Transfer-Encoding: quoted-printable

Dear Rats users:

I am writting a program to compute conditional forecasts of ARIMA =
models. To
do so I need to compute the ratio of the MA and AR polinomials, but I
haven=B4t been able find a command to do so. Tha SAS system has a =
command
called RATIO that does this. I suppose that there must be something =
similar
in RATS since it must use a multivariate version of it to compute =
impulse
response functions.

Can any one please help me?

In advance thanks a lot.

Best wishes,

Arturo Galindo

Research Department
Banco de la Rep=FAblica de Colombia


---------- End of message ----------

From: "Jian Yang" <jian-yang@tamu.edu>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: CATS equation and ECM residuals
Date: Fri, 12 Mar 1999 10:14:00 -0600
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
X-Mailer: Novell GroupWise 5.5 (via Mercury MTS (Bindery) v1.40)
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: quoted-printable

Dear RATS users,

I need your help to find how to save or retrieve the residuals for each =
equation in the system and ECM residuals after running CATS I(1) analysis. =
 Before, a RATS user briefly mentioned existence of these residuals and =
the way to save the ECM residuals. But I still have no clue. I consulted a =
professor who has used CATS almost since its inception, but he did not =
know where CATS saved these series.  There is a list of reserved names. =
Can someone tell me under which reserved name I can get the residual =
series or otherwise how I can save these residuals series in nonbatch/inter=
active mode?

Thank you for your time and help.


J. Yang
Ph.D. candidate
Texas A&M University
Colleg Station, Texas 77840


---------- End of message ----------

From: KHL2288@UTARLG.UTA.EDU
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: tests for sign and size bias effects
Date: Fri, 12 Mar 1999 20:52:26 -0600 (CST)
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
MIME-version: 1.0
Content-type: TEXT/PLAIN; CHARSET=US-ASCII
X-Mailer: Mercury MTS (Bindery) v1.40

Hi, RATS users.
Could anyone give me some references to interpret the tests for sign and size bias effects in GARCH.SRC.

Thanks a lot.


Kyong H. Lee
Dept. of Economics
University of Texas at Arlington

---------- End of message ----------

From: "Gregory, Richard" <RGregory@VEDP.state.va.us>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: RE: tests for sign and size bias effects
Date: Mon, 15 Mar 1999 08:19:55 -0500
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2448.0) (via Mercury MTS (Bindery) v1.40)
Content-Type: text/plain;

A good place to start is: "Engle, Robert F. and Victor K. Ng (1993)
"Measuring and Testing the Impact of News on Volatility", Journal of Finance
48, pp.1749-1778.

-----Original Message-----
From: KHL2288@UTARLG.UTA.EDU [mailto:KHL2288@UTARLG.UTA.EDU]
Sent: Friday, March 12, 1999 9:52 PM
To: RATS Discussion List
Subject: tests for sign and size bias effects


Hi, RATS users.
Could anyone give me some references to interpret the tests for sign and
size bias effects in GARCH.SRC.

Thanks a lot.


Kyong H. Lee
Dept. of Economics
University of Texas at Arlington

---------- End of message ----------

From: Irida Achthoven <achthovi@qed.econ.queensu.ca>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Maximum Likelihood Value
Date: Mon, 15 Mar 1999 10:18:09 -0800
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
Organization: Queen's University
X-Mailer: Mozilla 3.04 (Win16; I) (via Mercury MTS (Bindery) v1.40)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Dear Rats User's
I have to conduct a LR test on some simple garch models
I would like to know if some of you could tell me how to obtain the max.
Likelihood Value (also referred to as the LOG- L) using the garchuv prg.

I would really appreciate some help on that.
thank You
best regards,
Irida

---------- End of message ----------

From: Nesmith_T <Nesmith_T@bls.gov>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: DLL's
Date: Mon, 15 Mar 1999 14:15:01 -0500
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2232.9) (via Mercury MTS (Bindery) v1.40)
Content-Type: multipart/mixed;

This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.

------_=_NextPart_000_01BE6F18.2131CAD0
Content-Type: text/plain

Dear RATS Users,

Does RATS have the ability to invoke DLL's compiled in other languages?  I
believe the answer is no.  If that is correct, I believe such a feature
would be a useful addition to the program.  This would allow RATS
programmers to use the numerous libraries of numerical routines, developed
in FORTRAN and C for example, in their analysis.  Often the algorithms
programmed in those languages are significantly faster than a comparable
algorithm programmed in a statistical package.  While I have found RATS to
generally be pretty efficient, I would still expect to see some improvement.
Several other econometricians, with whom I work, have been able to decrease
the computation time for their simulations dramatically by utilizing DLLs,
written in C and called by their statistical package, to do the
computations.  I wondered whether this capability was planned for RATS and
whether their was interest in such a capability among the users?

Thanks,

Travis

- Travis D. Nesmith
   Bureau of Labor Statistics
   Division of Price and Index Number Research
   2 Massachusetts Ave., NE
   Room 3105
   Washington, DC  20212

   Voice (202) 606-6579 extension 597
   Fax    (202) 606-6853




------_=_NextPart_000_01BE6F18.2131CAD0
Content-Type: application/ms-tnef
Content-Transfer-Encoding: base64
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------_=_NextPart_000_01BE6F18.2131CAD0--

---------- End of message ----------

From: "mcymswa" <mcymswa@fs1.sm.umist.ac.uk>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Re: Maximum Likelihood Value
Date: Tue, 16 Mar 1999 12:33:46 BST
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
MIME-Version: 1.0
Content-type: text/plain; charset=US-ASCII
Content-transfer-encoding: 7BIT
X-mailer: Pegasus Mail for Windows (v2.54) (via Mercury MTS (Bindery) v1.40)

> From:          Irida Achthoven <achthovi@qed.econ.queensu.ca>
> To:            "RATS Discussion List" <RATS-L@efs.mq.edu.au>
> Subject:       Maximum Likelihood Value
> Date:          Mon, 15 Mar 1999 10:18:09 -0800
> Reply-to:      "RATS Discussion List" <RATS-L@efs.mq.edu.au>
> Organization:  Queen's University

> Dear Rats User's
> I have to conduct a LR test on some simple garch models
> I would like to know if some of you could tell me how to obtain the max.
> Likelihood Value (also referred to as the LOG- L) using the garchuv prg.
> 
> I would really appreciate some help on that.
> thank You
> best regards,
> Irida
>

It is printed as Function Value, before estimated parameters.

---------- End of message ----------

From: "Wolfgang Bauer" <Wolfgang_Bauer@gmx.net>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Present date
Date: Wed, 17 Mar 1999 19:32:00 +0100
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
MIME-Version: 1.0
Content-type: text/plain; charset=US-ASCII
Content-transfer-encoding: 7BIT
X-mailer: Pegasus Mail for Win32 (v3.01b) (via Mercury MTS (Bindery) v1.40)

Dear all,

is there a way to access the present date in RATS?

Thanks for any hints, best regards,

W. Bauer

---------- End of message ----------

From: jtebeka@oddo.fr
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: choice of law in garch models
Date: Tue, 23 Mar 1999 15:40:47 +0100
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
Mime-Version: 1.0
Content-type: text/plain; charset=us-ascii
X-Mailer: Mercury MTS (Bindery) v1.40




I am using the Garch procedure and i would like to know how i can test if a
student or ged is the appropriate distribution to use.
Thanks for your help.
Jacques



---------- End of message ----------

From: Christopher F Baum <baum@bc.edu>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Re: choice of law in garch models
Date: Tue, 23 Mar 1999 10:02:36 -0500
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
X-Mailer: Mulberry (MacOS) [1.4.2, s/n P020-300786-009] (via Mercury MTS (Bindery) v1.40)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Since student t (and normal) are special cases of GED, why wouldn't a test
for the est. thickness parameter be appropriate? If it is significantly
greater than 2 (its value for a normal), then a t-dist with thicker tails
would certainly be rejected in favor of a GED with tails thinner than the
normal.

Kit Baum
Boston College Economics

--On Tue, Mar 23, 1999 15:40 +0100 jtebeka@oddo.fr wrote:

> 
> 
> 
> I am using the Garch procedure and i would like to know how i can test if
> a student or ged is the appropriate distribution to use.
> Thanks for your help.
> Jacques
> 
> 





---------- End of message ----------

From: jtebeka@oddo.fr
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: =?iso-8859-1?Q?R=E9f._:_Re:_choice_of_law_in_garch_models?=
Date: Tue, 23 Mar 1999 16:22:32 +0100
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
Mime-Version: 1.0
Content-type: text/plain; charset=iso-8859-1
Content-transfer-encoding: quoted-printable
X-Mailer: Mercury MTS (Bindery) v1.40




do you mean that if d is (statiscally) greater than 2 : use a student
                            if d is lower than 2 : use a GED
             else use a normal
Moreover i would like to know if i have choose a distribution (student,=
 Ged
or Normal), how can i calculate the probability that X>value where X
follows student, Ged or Normal distribution
Thanks.
Jacques




Christopher F Baum <baum@bc.edu> le 23/03/99 16:02:36

Veuillez r=E9pondre =E0 "RATS Discussion List" <RATS-L@efs.mq.edu.au>

Pour :    "RATS Discussion List" <RATS-L@efs.mq.edu.au>
cc :   (ccc : Jacques Tebeka/ODDO)
Objet :   Re: choice of law in garch models


=

Content-type: text/plain; charset=us-ascii


Since student t (and normal) are special cases of GED, why wouldn't a test
for the est. thickness parameter be appropriate? If it is significantly
greater than 2 (its value for a normal), then a t-dist with thicker tails
would certainly be rejected in favor of a GED with tails thinner than the
normal.
Kit Baum
Boston College Economics
--On Tue, Mar 23, 1999 15:40 +0100 jtebeka@oddo.fr wrote:
>
>
>
> I am using the Garch procedure and i would like to know how i can test if
> a student or ged is the appropriate distribution to use.
> Thanks for your help.
> Jacques
>
>










---------- End of message ----------

From: Christopher F Baum <baum@bc.edu>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: =?ISO-8859-1?Q?Re:_R=E9f._:_Re:_choice_of_law_in_garch_models?=
Date: Tue, 23 Mar 1999 10:47:59 -0500
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
X-Mailer: Mulberry (MacOS) [1.4.2, s/n P020-300786-009] (via Mercury MTS (Bindery) v1.40)
MIME-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: quoted-printable

Nelson (econometrica 59:2,p.353) states, with respect to f(z), that 'v is =
a
tail thickness param. When v=3D2, z has std norm distr. For v<2, dist of z
has thicker tails than normal (e.g. when v=3D1, z has double exp dist) and
for v>2, the dist of z has thinner tails than the normal (e.g. for =
v=3Dinf, z
is uniformly dist on interval [-sqrt(3),sqrt(3)].'

So I think what you have written below is backwards (and perhaps my
original message implied that).

Eval of Pr[x>value] for normal or student t is standard eval of RATS
functions. 
Not sure how you would eval the GED cdf without writing it out.

Kit Baum


--On Tue, Mar 23, 1999 16:22 +0100 jtebeka@oddo.fr wrote:

> 
> 
> 
> do you mean that if d is (statiscally) greater than 2 : use a student
>                             if d is lower than 2 : use a GED
>              else use a normal
> Moreover i would like to know if i have choose a distribution (student,
> Ged or Normal), how can i calculate the probability that X>value where X
> follows student, Ged or Normal distribution
> Thanks.
> Jacques
> 
> 
> 
> 
> Christopher F Baum <baum@bc.edu> le 23/03/99 16:02:36
> 
> Veuillez r=E9pondre =E0 "RATS Discussion List" <RATS-L@efs.mq.edu.au>
> 
> Pour :    "RATS Discussion List" <RATS-L@efs.mq.edu.au>
> cc :   (ccc : Jacques Tebeka/ODDO)
> Objet :   Re: choice of law in garch models
> 
> 
> 
> Content-type: text/plain; charset





---------- End of message ----------

From: "Suzanna De Boef" <sdeboef@latte.harvard.edu>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: VAR.SRC
Date: Thu, 25 Mar 1999 11:18:46 -0500
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
X-Mailer: Mozilla 4.05 [en] (Win95; I) (via Mercury MTS (Bindery) v1.40)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

I have downloaded the Jan 12, 1999 version of VAR.SRC from Estima's
homepage and saved it to my c:\winrats\ directory (Rats version 4.30).
When I try to source the code using
    source c:\winrats\var.src

I receive the following error message:
    (01.1426)
(01.1426) COMPUTE lags = %rows(COEFFSBYLAG)-(NEXOG+DET-1)
(01.1460) COMPUTE k = lag = lags
<<<<
## SX27. Illegal Combination of Data Types for Operation

I have never received error messages simply by running the line calling
the source code before. Do I need to source this differently?

Also, the source code does not include an example, but includes the
following:
    VAR start end BO

Can someone tell me what  BO refers to?

Thank you.
Suzanna

--
Suzanna De Boef                                            Voice:
617-496-2272
Visiting Assistant Professor                               Fax:
617-496-2254
Center for Basic Research in the Social Sciences
Harvard University
34 Kirkland Street
Cambridge, MA 02138

http://www.la.psu.edu/~sdeboef/



---------- End of message ----------

From: Ying-Man Chung <y_mchung@yahoo.com>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: re:FM-ols
Date: Thu, 25 Mar 1999 09:09:54 -0800 (PST)
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
X-Mailer: Mercury MTS (Bindery) v1.40

Dear RATS users,


May I ask whether RATS procedures can estimate a regression model by use
of Fully-modified Ols, Give and GMM, plesase?

Thank you for your help!!!

With best wishes,


Sarah,

HKSYC 

_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com


---------- End of message ----------

From: "Dr. Sean Michael Snaith" <ssnaith@uwimona.edu.jm>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Re: VAR.SRC
Date: Thu, 25 Mar 1999 12:56:27 -0500
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
X-Mailer: QUALCOMM Windows Eudora Pro Version 4.0 (via Mercury MTS (Bindery) v1.40)
Mime-Version: 1.0
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable

Hi Suzanne,

I have had no problems compiling the VAR source program.

The BO is a user inputted method of orthogonalization. This would be guided=
 by
economic (political) theory of the underlying model.  Otherwise you could=
 use
one of the optional methods of orthogonalization within the  program
(Choleski,
Bernanke, etc).

If you are not sure about this, contact me directly.

Maybe Ray Lombra could give you a grant to come to Jamaica to finish this
research.

Cheers,

Sean







Dr. Sean Michael Snaith
Department of Economics
University of the West Indies
Kingston, 7
Jamaica
West Indies

TEL: (876) 977-1188=A0=20
FAX: (876) 977-1483


---------- End of message ----------

From: Norman Morin <m1njm00@frb.gov>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Re: VAR.SRC
Date: Thu, 25 Mar 1999 13:11:08 -0500
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
MIME-version: 1.0
X-Mailer: exmh version 2.0.2.4 7/8/98 (via Mercury MTS (Bindery) v1.40)
Content-type: text/plain; charset=us-ascii


> I receive the following error message:
>     (01.1426)
> (01.1426) COMPUTE lags = %rows(COEFFSBYLAG)-(NEXOG+DET-1)
> (01.1460) COMPUTE k = lag = lags
> <<<<
> ## SX27. Illegal Combination of Data Types for Operation

Do you have a variable named lags, coeffsbylag, nexog, det, lag,
or lags floating around?  It might be assigned to a different data
type from what the program is expecting.  I'm often remiss in declaring
things as local variables.

> Also, the source code does not include an example, but includes the
> following:
>     VAR start end BO

The BO is an optional parameter -- the procedure runs fine without it --
for a user-supplied matrix for orthogonalizing the shocks in the
system.

Regards,

Norm  

-- 

                           Norman J. Morin 
                  nmorin@frb.gov or m1njm00@frb.gov         
          --------------------------------------------------
          Division of Research and Statistics * Mail Stop 82
           Board of Governors of the Federal Reserve System 
               Washington, D.C. 20551 * (202) 452-2476
          --------------------------------------------------



---------- End of message ----------

From: "Suzanna De Boef" <sdeboef@latte.harvard.edu>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Re: VAR.SRC
Date: Thu, 25 Mar 1999 13:47:05 -0500
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
X-Mailer: Mozilla 4.05 [en] (Win95; I) (via Mercury MTS (Bindery) v1.40)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Hi Sean,

What a delightful thought, a semester or 2 in Jamaica! You obviously have Penn
State roots, I'd be curious to hear more.

Would you be willing to send a sample program file? Perhaps I can identify
something simple that is causing my difficulties. I am just starting a project
that tries to identify the sources of citizen economic evaluations as a function
of the real economy, media coverage of the economy, politics and policy. I've done
some VAR before, but mostly in the context of testing for Granger causality, so I
am excited to get the source code to work and see what we have so far.

I certainly appreciate your help.
Suzanna

Dr. Sean Michael Snaith wrote:

> Hi Suzanne,
>
> I have had no problems compiling the VAR source program.
>
> The BO is a user inputted method of orthogonalization. This would be guided by
> economic (political) theory of the underlying model.  Otherwise you could use
> one of the optional methods of orthogonalization within the  program
> (Choleski,
> Bernanke, etc).
>
> If you are not sure about this, contact me directly.
>
> Maybe Ray Lombra could give you a grant to come to Jamaica to finish this
> research.
>
> Cheers,
>
> Sean
>
> Dr. Sean Michael Snaith
> Department of Economics
> University of the West Indies
> Kingston, 7
> Jamaica
> West Indies
>
> TEL: (876) 977-1188
> FAX: (876) 977-1483



--
Suzanna De Boef                                            Voice: 617-496-2272
Visiting Assistant Professor                               Fax: 617-496-2254
Center for Basic Research in the Social Sciences
Harvard University
34 Kirkland Street
Cambridge, MA 02138

http://www.la.psu.edu/~sdeboef/



---------- End of message ----------

From: "Suzanna De Boef" <sdeboef@latte.harvard.edu>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Re: VAR.SRC
Date: Thu, 25 Mar 1999 13:56:02 -0500
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
X-Mailer: Mozilla 4.05 [en] (Win95; I) (via Mercury MTS (Bindery) v1.40)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Thanks so much, that was precisely it -- Lag for Index of Lagging
Indicators! It compiles now.
Suzanna

Norman Morin wrote:

> > I receive the following error message:
> >     (01.1426)
> > (01.1426) COMPUTE lags = %rows(COEFFSBYLAG)-(NEXOG+DET-1)
> > (01.1460) COMPUTE k = lag = lags
> > <<<<
> > ## SX27. Illegal Combination of Data Types for Operation
>
> Do you have a variable named lags, coeffsbylag, nexog, det, lag,
> or lags floating around?  It might be assigned to a different data
> type from what the program is expecting.  I'm often remiss in declaring
> things as local variables.
>
> > Also, the source code does not include an example, but includes the
> > following:
> >     VAR start end BO
>
> The BO is an optional parameter -- the procedure runs fine without it --
> for a user-supplied matrix for orthogonalizing the shocks in the
> system.
>
> Regards,
>
> Norm
>
> --
>
>                            Norman J. Morin
>                   nmorin@frb.gov or m1njm00@frb.gov
>           --------------------------------------------------
>           Division of Research and Statistics * Mail Stop 82
>            Board of Governors of the Federal Reserve System
>                Washington, D.C. 20551 * (202) 452-2476
>           --------------------------------------------------



--
Suzanna De Boef                                            Voice:
617-496-2272
Visiting Assistant Professor                               Fax: 617-496-2254

Center for Basic Research in the Social Sciences
Harvard University
34 Kirkland Street
Cambridge, MA 02138

http://www.la.psu.edu/~sdeboef/



---------- End of message ----------

From: "Suzanna De Boef" <sdeboef@latte.harvard.edu>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Re: VAR.SRC
Date: Thu, 25 Mar 1999 13:57:25 -0500
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
X-Mailer: Mozilla 4.05 [en] (Win95; I) (via Mercury MTS (Bindery) v1.40)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Sorry to bother you again, do you have a sample program file you might be
willing to share that uses VAR.SRC?
Thanks again,
suzanna

Norman Morin wrote:

> > I receive the following error message:
> >     (01.1426)
> > (01.1426) COMPUTE lags = %rows(COEFFSBYLAG)-(NEXOG+DET-1)
> > (01.1460) COMPUTE k = lag = lags
> > <<<<
> > ## SX27. Illegal Combination of Data Types for Operation
>
> Do you have a variable named lags, coeffsbylag, nexog, det, lag,
> or lags floating around?  It might be assigned to a different data
> type from what the program is expecting.  I'm often remiss in declaring
> things as local variables.
>
> > Also, the source code does not include an example, but includes the
> > following:
> >     VAR start end BO
>
> The BO is an optional parameter -- the procedure runs fine without it --
> for a user-supplied matrix for orthogonalizing the shocks in the
> system.
>
> Regards,
>
> Norm
>
> --
>
>                            Norman J. Morin
>                   nmorin@frb.gov or m1njm00@frb.gov
>           --------------------------------------------------
>           Division of Research and Statistics * Mail Stop 82
>            Board of Governors of the Federal Reserve System
>                Washington, D.C. 20551 * (202) 452-2476
>           --------------------------------------------------



--
Suzanna De Boef                                            Voice:
617-496-2272
Visiting Assistant Professor                               Fax: 617-496-2254

Center for Basic Research in the Social Sciences
Harvard University
34 Kirkland Street
Cambridge, MA 02138

http://www.la.psu.edu/~sdeboef/



---------- End of message ----------

From: Moissinac <moissinac@asp.ifw.uni-kiel.de>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Simulating Impulse-Response with an extension to a VAR
Date: Fri, 26 Mar 1999 10:07:39 +0100
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
Organization: ASP International Policy, IfW
X-Mailer: Mozilla 4.01 [en] (Win95; I) (via Mercury MTS (Bindery) v1.40)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Hi everybody,

I have a problem simulating some impulse response function with a
linear system that includes a Structural VAR (7 equations) and 5 other
equations which I construct (these 5 new variables are just particular
combination of the 7 variables of the VAR).

Moreover, I also extend the structural scheme previously estimated for
the
VAR. So I have 7+5 linear autoregressive equations, and a (14,14) BMAT
and a (14,14) FACTOR. (I use var.src)

I thought I had just to use the following instructions:

system 1 to 14
end(system)

and then run an impulse.src with the FACTOR just constructed.

But it does not give coherent impulse -reponse functions: in fact,
the last 5 equations do not show any dynamics!

It would be terrific if someone could help me,

Thanks,

Vincent Moissinac


---------- End of message ----------

From: "Wolfgang Bauer" <Wolfgang_Bauer@gmx.net>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: End programme in if branch
Date: Fri, 26 Mar 1999 11:32:11 +0100
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
MIME-Version: 1.0
Content-type: text/plain; charset=US-ASCII
Content-transfer-encoding: 7BIT
X-mailer: Pegasus Mail for Win32 (v3.01b) (via Mercury MTS (Bindery) v1.40)

Dear all,

in a procedure, I would like to stop the whole programme if a certain
condition is fulfilled. IF CONDITION 
 {...
 HALT
 }

end the whole RATS what I not intend. 

IF CONDITION
 {...
 END
 }

on the other hand does not work. What is the right way to do this?

Best regards

W. Bauer



---------- End of message ----------

From: Michael Hanson <mhanson@econ.lsa.umich.edu>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Re: Simulating Impulse-Response with an extension to a VAR
Date: Fri, 26 Mar 1999 11:20:31 -0500 (EST)
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
X-Mailer: Mercury MTS (Bindery) v1.40

	I haven't used VAR.SRC, so you may need to modify this advice
somewhat.  I've coded something similar a while back, so this is from
memory....

	First, you need to create "accounting identities" which link the
other variables to the ones you estimated, using EQU(IDENTITY) and
ASSOCIATE.  It's not enough to only define the variables at the beginning
of your program. 

	Second, you need to include these identities in the IMPULSE
command:  in your case, the first 7 cards would be the estimated
variables, and the last 5 would be the identities.  (There are, of course,
only 7 "shocks", not 12.) 

	Since this is likely "automated" in VAR.SRC, you may need either
to edit the code or to write the program yourself.  I'm afraid others will
have to help you there.  Good luck.

					-- Mike

======================================================================
Michael S. Hanson                                    mhanson@umich.edu
Department of Economics         http://www.econ.lsa.umich.edu/~mhanson
University of Michigan


On Fri, 26 Mar 1999, Moissinac wrote:

> Hi everybody,
> 
> I have a problem simulating some impulse response function with a
> linear system that includes a Structural VAR (7 equations) and 5 other
> equations which I construct (these 5 new variables are just particular
> combination of the 7 variables of the VAR).
> 
> Moreover, I also extend the structural scheme previously estimated for
> the
> VAR. So I have 7+5 linear autoregressive equations, and a (14,14) BMAT
> and a (14,14) FACTOR. (I use var.src)
> 
> I thought I had just to use the following instructions:
> 
> system 1 to 14
> end(system)
> 
> and then run an impulse.src with the FACTOR just constructed.
> 
> But it does not give coherent impulse -reponse functions: in fact,
> the last 5 equations do not show any dynamics!
> 
> It would be terrific if someone could help me,
> 
> Thanks,
> 
> Vincent Moissinac
> 
> 


---------- End of message ----------

From: Moissinac <moissinac@asp.ifw.uni-kiel.de>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Re: Simulating Impulse-Response with an extension to a VAR
Date: Fri, 26 Mar 1999 19:40:07 +0100
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
Organization: ASP International Policy, IfW
X-Mailer: Mozilla 4.01 [en] (Win95; I) (via Mercury MTS (Bindery) v1.40)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Thank you very much for your email, I see now at least two
of my mistakes. I correct my program (I use var.src only for estimating
the
preliminary VAR, and only then I include it into a larger system)
and I hope the results will be more coherent.

Once again, thank you,
Sincerely,

Vincent Moissinac




Michael Hanson wrote:

>         I haven't used VAR.SRC, so you may need to modify this advice
> somewhat.  I've coded something similar a while back, so this is from
> memory....
>
>         First, you need to create "accounting identities" which link
> the
> other variables to the ones you estimated, using EQU(IDENTITY) and
> ASSOCIATE.  It's not enough to only define the variables at the
> beginning
> of your program.
>
>         Second, you need to include these identities in the IMPULSE
> command:  in your case, the first 7 cards would be the estimated
> variables, and the last 5 would be the identities.  (There are, of
> course,
> only 7 "shocks", not 12.)
>
>         Since this is likely "automated" in VAR.SRC, you may need
> either
> to edit the code or to write the program yourself.  I'm afraid others
> will
> have to help you there.  Good luck.
>
>                                         -- Mike
>
> ======================================================================
>
> Michael S. Hanson                                    mhanson@umich.edu
>
> Department of Economics         http://www.econ.lsa.umich.edu/~mhanson
>
> University of Michigan
>
> On Fri, 26 Mar 1999, Moissinac wrote:
>
> > Hi everybody,
> >
> > I have a problem simulating some impulse response function with a
> > linear system that includes a Structural VAR (7 equations) and 5
> other
> > equations which I construct (these 5 new variables are just
> particular
> > combination of the 7 variables of the VAR).
> >
> > Moreover, I also extend the structural scheme previously estimated
> for
> > the
> > VAR. So I have 7+5 linear autoregressive equations, and a (14,14)
> BMAT
> > and a (14,14) FACTOR. (I use var.src)
> >
> > I thought I had just to use the following instructions:
> >
> > system 1 to 14
> > end(system)
> >
> > and then run an impulse.src with the FACTOR just constructed.
> >
> > But it does not give coherent impulse -reponse functions: in fact,
> > the last 5 equations do not show any dynamics!
> >
> > It would be terrific if someone could help me,
> >
> > Thanks,
> >
> > Vincent Moissinac
> >
> >




---------- End of message ----------

From: Liz Turner <elizabeth.turner@strath.ac.uk>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: 
Date: Fri, 26 Mar 1999 19:57:23 +0000
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
X-Mailer: QUALCOMM Windows Eudora Light Version 3.0.3 (32) (via Mercury MTS (Bindery) v1.40)
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"

Dear all, 

I am trying to do dynamic panel data estimation using RATS and was hoping
somebody out there might have a program to test for serial correlation or
perhaps some advice on the most appropriate test to use.

Any suggestions would be most appreciated.

Liz Turner



~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Liz Turner, University of Strathclyde, Department of Economics,
Curran Building, 100 Cathedral Street, Glasgow, United Kingdom,
G4 OLN

Tel: +44-(0)141-548-3859
Fax: +44-(0)141-552-5589
E mail: elizabeth.turner@strath.ac.uk
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

---------- End of message ----------

From: Price SG <S.G.Price@city.ac.uk>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: Re: your mail
Date: Sat, 27 Mar 1999 13:30:57 +0000 (BST)
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
X-Mailer: Mercury MTS (Bindery) v1.40

> I am trying to do dynamic panel data estimation using RATS and was hoping
> somebody out there might have a program to test for serial correlation or
> perhaps some advice on the most appropriate test to use.
> 

What does your dynamic panel look like?  If it's low T high N you are
probably better off using the Arellano and Bond method with their
software; if it's roughly same size T and N you should be doing a
Pesaran-Shin-Smith pooled mean group estimation in which case the standard
equation diagnostics are OK.  Apologies if I am suggesting things you are
already on top of.
  



*******************************************************************
 Prof Simon Price         email:      s.g.price@city.ac.uk 
 Department of Economics
 City University          Tel (work): +44 (0)171 477 8000 ext 4503
 Northampton Square       Fax (work): +44 (0)171 477 8580
 London EC1V 0HB UK       
 URL http://www.city.ac.uk/~sm344
*******************************************************************




---------- End of message ----------

From: Karin Elsner <elsner@iamo.uni-halle.de>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: saving data in .wk3 format
Date: Tue, 30 Mar 1999 16:54:28 +0200
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
Organization: IAMO
X-Mailer: Mozilla 3.01 [de] (Win95; I) (via Mercury MTS (Bindery) v1.40)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Dear RATS-users,
I want to include / save  some series that have been generated within
RATS procedures into an already existing data file in .wk3 format. I
only see how it works for RATS format files. What would you recommend?
Thank you!
Karin Elsner
-- 
*********************************************************************
                                                                      
  Karin Elsner             Institute of Agricultural Development      
                           in Central and Eastern Europe (IAMO)       
                           Emil-Abderhalden-Strasse 20                
                           06108 Halle                                
                           Germany                                    
                                                                      
                           --------------------------------------     
                                                                      
                           Phone +49 345 / 5 52 29 33                 
                           Fax   +49 345 / 5 00 81 77                 
                           Email elsner@iamo.uni-halle.de                
                    http://www.landw.uni-halle.de/iamo/iamo_e.htm     
                                                                      
*********************************************************************

---------- End of message ----------

From: Irida Achthoven <achthovi@qed.econ.queensu.ca>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: one step ahead  GARCH forecast (out of sample)
Date: Tue, 30 Mar 1999 17:32:51 -0800
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
Organization: Queen's University
X-Mailer: Mozilla 3.04 (Win16; I) (via Mercury MTS (Bindery) v1.40)
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Dear Rats user's,
Can somebody help me with a sample of one step ahead forecast for a
GARCH(1,1) model.
I have to forecast GARCH(1,1), with varous conditional mean models, RW,
AR(1), MA(1). I have to add each forecast from each step to the sample,
hence the sample size grows as we go along. and I have to forecast for 2
years. 
if any body can help me with a sample I will really appreciate it.
Best regards,
Irida

---------- End of message ----------

From: Joseph Malm /Svetlana Kutuzova-Malm <jmalm@unm.edu>
To: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Subject: math fonts???
Date: Tue, 30 Mar 1999 22:22:55 -0700
Errors-to: <rats-l-owner@efs.mq.edu.au>
Reply-to: "RATS Discussion List" <RATS-L@efs.mq.edu.au>
Sender: Maiser@efs1.efs.mq.edu.au
X-listname: <RATS-L@efs1.efs.mq.edu.au>
X-Mailer: Mozilla 4.05 [en] (Win95; I) (via Mercury MTS (Bindery) v1.40)
MIME-Version: 1.0
Content-Type: text/plain; charset=koi8-r
Content-Transfer-Encoding: 7bit

Dear List, A bit off the usual subject matter but does anyone know where I
can download (or can anyone email me one) ttf math symbol fonts? thanks
Joseph Malm
jmalm@unm.edu




---------- End of message ----------