I'm trying to run the following regressions uising the dataset attatched.
Share Price Change = Market Change + Dummy (WIN) + Dummy (lose) + Dummy (DRAW) looking at Share Price changes in relation to a result.
But each time I try I get the 'Near Singular Matrix' error message????????? what can i do to rectify?
Also Whats the difference between dropping the 'c' in the regression and dropping a variable e.g. Dummy (Draw)???
Any help would be much appreciated
Please find dataset attatched
Kind Regards
help
Moderators: EViews Gareth, EViews Moderator
help
- Attachments
-
- effort1.WF1
- evies
- (21.98 KiB) Downloaded 275 times
-
- Non-normality and collinearity are NOT problems!
- Posts: 3775
- Joined: Wed Sep 17, 2008 2:25 pm
Re: help
dunk4lcfc wrote:I'm trying to run the following regressions uising the dataset attatched.
Share Price Change = Market Change + Dummy (WIN) + Dummy (lose) + Dummy (DRAW) looking at Share Price changes in relation to a result.
But each time I try I get the 'Near Singular Matrix' error message????????? what can i do to rectify?
Also Whats the difference between dropping the 'c' in the regression and dropping a variable e.g. Dummy (Draw)???
Any help would be much appreciated
Please find dataset attatched
Kind Regards
You probably included a constant as well as all the dummies. One needs to be dropped. Which one is dropped doesn't matter, except for the interpretation of the coefficients. The one that is dropped becomes the "reference group."
Re: help
OK! But if I'm looking for say.... the effect a Win has on the Share Price change
the effect a Draw has on the Share Price Change
the effect a Defeat has on the Share Price Change
how am I going to be assured that the values I have are correct. As I have run seperate regressions, firstly leaving out 'C', then 'lose', then 'draw'.. but I get different values for each variable?? which one do I use, when writing up the results??
sorry for being soooo thick on this.
the effect a Draw has on the Share Price Change
the effect a Defeat has on the Share Price Change
how am I going to be assured that the values I have are correct. As I have run seperate regressions, firstly leaving out 'C', then 'lose', then 'draw'.. but I get different values for each variable?? which one do I use, when writing up the results??
sorry for being soooo thick on this.
-
- Non-normality and collinearity are NOT problems!
- Posts: 3775
- Joined: Wed Sep 17, 2008 2:25 pm
Re: help
dunk4lcfc wrote:OK! But if I'm looking for say.... the effect a Win has on the Share Price change
the effect a Draw has on the Share Price Change
the effect a Defeat has on the Share Price Change
how am I going to be assured that the values I have are correct. As I have run seperate regressions, firstly leaving out 'C', then 'lose', then 'draw'.. but I get different values for each variable?? which one do I use, when writing up the results??
sorry for being soooo thick on this.
As an experiment, save the forecast values from each regression. You'll see they are all the same.
Re: help
yes... I see that whichever variable I miss out then the 'c' value takes that same value as was when I regressed SP Change Mkt Change WIN LOSE DRAW
But Analytically.... how can it be described?
i.e by running the regression without the 'WIN' variable, then 'c' takes the value 0.006906. If it run the regression without 'C' then the win variable is 0.006906
But, when I run other regressions, namely dropping the 'Lose' and 'Draw' variables... I get a different 'WIN' coefficient then?? Whats its interpretation?
and in re. to my earlier post i.e. 'describing the change a win has on share price'...... which value do I use
Kind Regards
But Analytically.... how can it be described?
i.e by running the regression without the 'WIN' variable, then 'c' takes the value 0.006906. If it run the regression without 'C' then the win variable is 0.006906
But, when I run other regressions, namely dropping the 'Lose' and 'Draw' variables... I get a different 'WIN' coefficient then?? Whats its interpretation?
and in re. to my earlier post i.e. 'describing the change a win has on share price'...... which value do I use
Kind Regards
Return to “Econometric Discussions”
Who is online
Users browsing this forum: No registered users and 27 guests