RMSE

[kojanvis]

Hi all dear. Pleas help me to calculate the root mean square error (RMSE) of the coefficients by STATA for LSDVC estimator.

[Jorge Eduardo Perez Perez]

It may help if you provide detailes on this "LSVDC" estimator. My intuition tells me this is a least square dummy variable estimator, but how hould I know that if I were not into econometrics? Statalist is intended for a broad audience.

Usually the RMSE is a model statistic, not a statistic associated with coefficients. In linear regression, this would be the root of sum of the squared residuals divided by the degrees of freedom.It would be the same for a least squares dummy variable estimator.

[Sebastian Kripfganz]

I guess you are referring to the standard errors and not the RMSE. Please read the help file of xtlsdvc:

Remarks

xtlsdvc does not report analytical standard errors. Only bootstrap standard
errors are reported, provided that vcov(#) is given.

Bootstrap standard errors are downward biased when values for the unknown
parameters are supplied through the matrix my, since the procedure, keeping
my fixed over replications, neglects a source of varibility of the
bias-corrected LSDV estimator.

[kojanvis]

Dear Sebastian and Jorge Eduardo thank for you advice . I have already red the help file. As you know the important criteria to compare the LSDVC estimator with other dynamic model estimators is the difference between mean square error of estimated coefficients. I have seen many papers that authors reffed to this subject such as 'Estimating dynamic panel models in corporate finance'
Mark J. Flannery a, Kristine Watson, Journal of Corporate Finance 19 (2013) 1–19 (page 5 and table 4) But I do not know how they measure it. It needs to have the real coefficient of studied public. but it is not available.
I have initially think like Jorge Eduardo but in mentioned paper they have measured RMSE separably for coefficient of lag depended variable and other explanatory variables. it is different from sum of the squared residuals divided by the degrees of freedom.

[kojanvis]

It may help if you provide detailes on this "LSVDC" estimator. My intuition tells me this is a least square dummy variable estimator, but how hould I know that if I were not into econometrics? Statalist is intended for a broad audience.

Usually the RMSE is a model statistic, not a statistic associated with coefficients. In linear regression, this would be the root of sum of the squared residuals divided by the degrees of freedom.It would be the same for a least squares dummy variable estimator.

Dear Sebastian and Jorge Eduardo thank for you advice . I have already red the help file. As you know the important criteria to compare the LSDVC estimator with other dynamic model estimators is the difference between mean square error of estimated coefficients. I have seen many papers that authors reffed to this subject such as 'Estimating dynamic panel models in corporate finance'
Mark J. Flannery a, Kristine Watson, Journal of Corporate Finance 19 (2013) 1–19 (page 5 and table 4) But I do not know how they measure it. It needs to have the real coefficient of studied public. but it is not available.
I have initially think like Jorge Eduardo but in mentioned paper they have measured RMSE separably for coefficient of lag depended variable and other explanatory variables. it is different from sum of the squared residuals divided by the degrees of freedom.

[kojanvis]

thanks a lot for all of you to help me about RMSE in LSDVC model.

[Carlo Lazzaro]

kojanvis (as per FAQ #6 please note the preference on this forum for real full names. Just click on the bottom-right Contact us button to re-register accordingly. Thanks):
I would e-mail the corresponding author of the article you quoted and ask for the details you're interested in.