the standardQmodels of investment byWhited (1992) and Hubbard and Kashyap (1992

Thanks Mary...I will explore it

The main problem you face is probable the low N. You can have a look at this paper

https://home.uchicago.edu/~amshaikh/webfiles/wild.pdf

Many models using this sort of bootstrap can be implemented with Stata.

Thank you Prof.Jose-Ignacio for your valuable suggestion. I shall look for it, Hope it will work well for our analysis.

Regards

P C Padhan

"Result is not coming good" could be quite valuable. It means that you should reject your model or its specification, respectively. That is how scientific research functions or should function: One has a hypothesis and one should try to test it. If the test speaks against it, the hypothesis is false and one may accept this proof or one could try to find an alternative hypothesis. I think it not feasible, to apply different methods (including dummies), until you get results that please you and publish only these results.

By the way, it is difficult or impossible to give advice for an unknown equation/specification.

Dear Prof. Anton

Thanks for your valuable comments and suggestion. My results accept the null hypothesis in the context of the study, which perhaps goes against the theory. It means the theory is not valid in the context of the study and time period. You are right that one should not try to get the result that please us, rather the method, purpose and steps of the empirical analysis should be rightly followed. I will relook it once again with more in-depth analysis, perhaps that might help me resolve the problem. Thanks a ton once again for your comments.

best

P C P

Dear Purna

Kindly recheck your model, methods of analysis and your variables.

You'r result may not fall in line with your theory sir.

That's why it's a Research.

Regards

Dear Bello

You'r right...i have been struggling to get the good result....The result is not consistent with theory. However if I sent the paper with this result, the reviewer might reject the paper. So we are looking for some better method. Looking for some more suggestions.

Best

P C Padhan

You can as well try the LSDVC or other bootstrap fixed effect dynamic model provided that your independent variable is strictly exogenous.. More so, you take the rogour of testing the stationarity of the variables to decide on the appropriate method CS-DL, CS-ARDL, panel var etc...

Thanks Hamid. I shall surely try for it.

regards

PCP