Pooled Regression
Moderators: EViews Gareth, EViews Moderator
Pooled Regression
I am trying to estimate a pooled regression using commands (rather than interactively). I have tried the following:
pool e_x1.ls ?x1 c ?y1
where a number of variables with the suffixes "x1" and "y1" exist in the workfile. This cresaes a pool pbject called e_x1, however, when I open it it contains the following:
Cross Section Identifiers: (Enter identifiers below this line)
X1?
C
Y1?
rather than the regression results as I had hoped. Any idea of what I am doing wrong?
pool e_x1.ls ?x1 c ?y1
where a number of variables with the suffixes "x1" and "y1" exist in the workfile. This cresaes a pool pbject called e_x1, however, when I open it it contains the following:
Cross Section Identifiers: (Enter identifiers below this line)
X1?
C
Y1?
rather than the regression results as I had hoped. Any idea of what I am doing wrong?
-
- EViews Developer
- Posts: 2672
- Joined: Wed Oct 15, 2008 9:17 am
Re: Pooled Regression
You're combining two commands. The pool statement is used to declare a pool object and to define the cross-section identifiers. The ls is an estimation command used with a specific pool object. So your command tells EViews to declare a pool object E_X1 and specify the identifiers "?x1" "c" and "?y1" (the .ls is ignored by the pool command), hence your results.
What you need to do is to first define the pool identifiers using one command, and then estimate with a second:
The second statement tells EViews to use the existing pool E_X1 to estimate a specification. Note in particular that we do not use the pool command in the second line.
What you need to do is to first define the pool identifiers using one command, and then estimate with a second:
Code: Select all
pool e_x1 [list of identifiers]
e_x1.ls ?x1 c ?y1
The second statement tells EViews to use the existing pool E_X1 to estimate a specification. Note in particular that we do not use the pool command in the second line.
Re: Pooled Regression
Hi
I am trying to run a pool regression on Eviews 5 that is described as:
DPRu = at+b1Current Ratio1t+b2 Liquidity Ratio2t +b3Gearing Ratio 3t +b4 Profit Margin Ratio4t +b5 Asset Cover5t + ut (in Response to yu= at+b1x1t+b2x2t+b3x3t+b4x4t+b5x5t+ut) with DPR being my dependent variable and the rest indipendent variables.
but it seems that I am not getting the results i was supposed to.
Can you please instruct me on the way i can run the regression correctly?
Thank you in advance
I am trying to run a pool regression on Eviews 5 that is described as:
DPRu = at+b1Current Ratio1t+b2 Liquidity Ratio2t +b3Gearing Ratio 3t +b4 Profit Margin Ratio4t +b5 Asset Cover5t + ut (in Response to yu= at+b1x1t+b2x2t+b3x3t+b4x4t+b5x5t+ut) with DPR being my dependent variable and the rest indipendent variables.
but it seems that I am not getting the results i was supposed to.
Can you please instruct me on the way i can run the regression correctly?
Thank you in advance
-
- Fe ddaethom, fe welon, fe amcangyfrifon
- Posts: 13317
- Joined: Tue Sep 16, 2008 5:38 pm
Re: Pooled Regression
Could you describe what you have done, and why it didn't result in the results you were expecting?
Follow us on Twitter @IHSEViews
Re: Pooled Regression
I imported stacked data into eviews 5 with Dpr (dependent variable) , current_ratio, profit_margin, asset_cover, quick_ratio, gearing (independent variables) in accordance with the equation DPRu = at+b1Current Ratio1t+b2 Liquidity Ratio2t ++b3Gearing Ratio 3t +b4 Profit Margin Ratio4t +b5 Asset Cover5t + ut
Then i created a pool object with all the 6 cross-section identifiers and then used pooled estimation identifying dependent variable and common coefficients.
However, all my results are significant with some extraordinary coefficients even tho in other studies there were some mixed results.
Is that right or am i doing something wrong?
Thank you in advance for your help
Then i created a pool object with all the 6 cross-section identifiers and then used pooled estimation identifying dependent variable and common coefficients.
However, all my results are significant with some extraordinary coefficients even tho in other studies there were some mixed results.
Is that right or am i doing something wrong?
Thank you in advance for your help
-
- Fe ddaethom, fe welon, fe amcangyfrifon
- Posts: 13317
- Joined: Tue Sep 16, 2008 5:38 pm
Re: Pooled Regression
should i generate a series by equation? or it's fine the way it is? because i thought maybe it's that
-
- Fe ddaethom, fe welon, fe amcangyfrifon
- Posts: 13317
- Joined: Tue Sep 16, 2008 5:38 pm
-
- Non-normality and collinearity are NOT problems!
- Posts: 3775
- Joined: Wed Sep 17, 2008 2:25 pm
Re: Pooled Regression
If you're really worried about results that are too good , you might post a picture of your estimation results and see if anyone has a suggestion.
Re: Pooled Regression
Dependent Variable: DPR
Method: Pooled Least Squares
Date: 11/23/10 Time: 20:15
Sample: 1 150
Included observations: 150
Cross-sections included: 6
Total pool (balanced) observations: 900
Variable Coefficient Std. Error t-Statistic Prob.
GEARING 0.046352 0.055954 0.828401 0.4077
QUICK_RATIO -196.5558 41.61238 -4.723494 0.0000
ASSET_COVER -0.008172 0.179668 -0.045483 0.9637
PROFIT_MARGIN -0.186320 0.630500 -0.295512 0.7677
CURRENT_RATIO 215.7294 35.46695 6.082548 0.0000
This is what i get from running a regression with 150 obs....is this correct?
R-squared -0.008351 Mean dependent var 94.76391
Adjusted R-squared -0.012857 S.D. dependent var 254.3747
S.E. of regression 256.0048 Akaike info criterion 13.93381
Sum squared resid 58656898 Schwarz criterion 13.96049
Log likelihood -6265.214 Durbin-Watson stat 2.070993
Method: Pooled Least Squares
Date: 11/23/10 Time: 20:15
Sample: 1 150
Included observations: 150
Cross-sections included: 6
Total pool (balanced) observations: 900
Variable Coefficient Std. Error t-Statistic Prob.
GEARING 0.046352 0.055954 0.828401 0.4077
QUICK_RATIO -196.5558 41.61238 -4.723494 0.0000
ASSET_COVER -0.008172 0.179668 -0.045483 0.9637
PROFIT_MARGIN -0.186320 0.630500 -0.295512 0.7677
CURRENT_RATIO 215.7294 35.46695 6.082548 0.0000
This is what i get from running a regression with 150 obs....is this correct?
R-squared -0.008351 Mean dependent var 94.76391
Adjusted R-squared -0.012857 S.D. dependent var 254.3747
S.E. of regression 256.0048 Akaike info criterion 13.93381
Sum squared resid 58656898 Schwarz criterion 13.96049
Log likelihood -6265.214 Durbin-Watson stat 2.070993
-
- Non-normality and collinearity are NOT problems!
- Posts: 3775
- Joined: Wed Sep 17, 2008 2:25 pm
Re: Pooled Regression
It looks like you're missing a constant term.
Re: Pooled Regression
Dependent Variable: DPR
Method: Pooled Least Squares
Date: 11/23/10 Time: 21:33
Sample: 1 150
Included observations: 150
Cross-sections included: 6
Total pool (balanced) observations: 900
Variable Coefficient Std. Error t-Statistic Prob.
C 129.7147 26.76230 4.846921 0.0000
GEARING -0.073342 0.060530 -1.211655 0.2260
CURRENT_RATIO 83.31521 44.42305 1.875495 0.0611
QUICK_RATIO -108.5845 44.92838 -2.416835 0.0159
PROFIT_MARGIN -1.632089 0.690477 -2.363711 0.0183
ASSET_COVER 0.116669 0.179312 0.650647 0.5154
R-squared 0.017468 Mean dependent var 94.76391
Adjusted R-squared 0.011973 S.D. dependent var 254.3747
S.E. of regression 252.8473 Akaike info criterion 13.91009
Sum squared resid 57154973 Schwarz criterion 13.94211
Log likelihood -6253.542 F-statistic 3.178886
Durbin-Watson stat 2.122701 Prob(F-statistic) 0.007497
is this how it should be?
Method: Pooled Least Squares
Date: 11/23/10 Time: 21:33
Sample: 1 150
Included observations: 150
Cross-sections included: 6
Total pool (balanced) observations: 900
Variable Coefficient Std. Error t-Statistic Prob.
C 129.7147 26.76230 4.846921 0.0000
GEARING -0.073342 0.060530 -1.211655 0.2260
CURRENT_RATIO 83.31521 44.42305 1.875495 0.0611
QUICK_RATIO -108.5845 44.92838 -2.416835 0.0159
PROFIT_MARGIN -1.632089 0.690477 -2.363711 0.0183
ASSET_COVER 0.116669 0.179312 0.650647 0.5154
R-squared 0.017468 Mean dependent var 94.76391
Adjusted R-squared 0.011973 S.D. dependent var 254.3747
S.E. of regression 252.8473 Akaike info criterion 13.91009
Sum squared resid 57154973 Schwarz criterion 13.94211
Log likelihood -6253.542 F-statistic 3.178886
Durbin-Watson stat 2.122701 Prob(F-statistic) 0.007497
is this how it should be?
-
- Non-normality and collinearity are NOT problems!
- Posts: 3775
- Joined: Wed Sep 17, 2008 2:25 pm
Re: Pooled Regression
That might be right, or you might want to allow for cross-section specific constants.
Re: Pooled Regression
I'll stick to what i've got so far. Thank you for your patience and your help
Re: Pooled Regression
sorry for bothering again but can you please tell me how can i estimate the significance of the variables i found? Is it the t-statistics i am using? the probability or the coefficient?
Who is online
Users browsing this forum: No registered users and 33 guests