arellano-bover/blundell-bond

[Emmanuel Owusu Oppong]

Hello people, I have a panel data of 9 firms for a period of 11 years with no missing data. In 2017, a new policy was introduced which affects all nine firms. I want to assess the policy on the nine firms so I have 7 explanatory variables including a dummy for when the policy was introduced.

Is it okay to use arellano-bover/blundell-bond model If my data has 9 firms and 11 years?

Any other model of econometrics book anyone can recommend on models

Thank you

[Sebastian Kripfganz]

These GMM methods for dynamic panel data models are designed for data sets where N (the number of firms) is large. With just 9 firms, I would not apply those methods. You would need to think about a conventional instrumental-variables estimation, or just a standard fixed-effects regression.

[Emmanuel Owusu Oppong]

These GMM methods for dynamic panel data models are designed for data sets where N (the number of firms) is large. With just 9 firms, I would not apply those methods. You would need to think about a conventional instrumental-variables estimation or just a standard fixed-effects regression.

Thanks, Sebastian. How about using the maximum likelihood model? Linear Dynamic Panel-data Estimation Using Maximum Likelihood and Structural Equation Modeling (sagepub.com)
With the standard fixed-effect model, any idea or recommendations on robust checking? My teacher said there should be a robust check after the main regression.
Thank you

[Sebastian Kripfganz]

How about using the maximum likelihood model? Linear Dynamic Panel-data Estimation Using Maximum Likelihood and Structural Equation Modeling (sagepub.com)

This approach also requires rather large data sets.

I cannot really tell what specific robustness checks might be required in your case. This might be a question to ask your teacher.

[Emmanuel Owusu Oppong]

This approach also requires rather large data sets.

I cannot really tell what specific robustness checks might be required in your case. This might be a question to ask your teacher.

Thanks so much for the reply. I will check with him to see