The RATS Software Forum

So for both Models, I put the dummy variable in the mean and the variance equation

I am trying to construct a dummy variable in both a AR(1) GJR and AR (1) Garch-m to measure the impact of a short selling ban on the leverage effect and risk. I.m not sure where to place the dummy variable to measure the impact correctly. Would I place the dummy in the variance equation for the GJR ...

You might want to try using t rather than Normal errors (DISTRIB=T) option, given the high kurtosis on the standardized residuals.

Hi Tom, When your using the T distribution option, under it in the shape box what do i put? Is it the degrees of freedom. my Obs are 1999

I'm not sure, thats why i asked. I thought for some reason all the coefficients in the model needed to be.

I am trying to capture the volatility in the stock returns and the dummy variable to show what the impact short selling ban
had on it

Cheers Tom, so the end result is theres no auto left in the residuals and it does a good job explaining the volatility. and if the P-value from the tests is > 0.5 there no auto left i.e the model is good. one other thing the MA2 p-value is insignificant, should this significant whats the interpretat...

Id be lost without you Tom really appreciate your help, I know my questions my are very basic but im trying to learn. I am trying to get the impact of the short selling ban on the volatility in stock returns. i used a MA 2 Garch 1,1 with a dummy both in the mean and variance. I am trying to interpre...

Hi Tom,
I checked out the manual after your advice.
for a MA2 for Garch (1,1) In the wizard, would I put into the Mean Model Variables section:
%mvgavge{2} or %mvgavge{1 to 2} im not sure which is correct

Thanks tom, sorry for being a pain, but my next problem is, i am trying to run the ARMA models of best fit for my return series in garch. i want to run ARMA (4,2) Garch ,ARMA (3,2) Garch and ARMA (1,1) Garch. i am not very sure how to fit in the ARMA model in to the mean equation with the wizard. Do...

Thanks for your reply Tom,

I took the dfc=%nreg out of the code from here cor(partial=pacf,qstats,number=12,span=4,dfc=%nreg) resids and reran it.
The Qstats completely changed. is this part of the code necessary for the procedure to run correctly

Hi, i am trying to find a model of best fit for daily stock % returns using the box jenkins method to test for volatility . My problem for all these models i am getting back no auto present with Ljung Box Qstat and for the Q4 is coming back blank for ARMA 1,2 AND ARMA 2,. All the Qstats are giving b...

So i didn't make an mistake in my code, and the end result is that there is arch effects present, so a garch model is required?

I used box jinkins to find the appropriate model for my stock returns. BJ suggested arma 1-1. I tried to test it for remain arch effects by using this code below. I am not sure im doing this correctly, has i also checked the 4 other, AR1 AR2 MA1 ARMA 1-1 ARMA 2-1 and 1-2 they all came back the same:...

Thanks Tom, I appreciate your advice.

Sorry in Advance!!! I am very new to RATS, never used before in fact. I am currently doing my thesis on the relationship between volatility and the introduction of a short selling ban on financial stocks in Ireland. I am using Garch models( trying to use) in RATS. My problem is that in my Econometri...