[Aryan Bose]

FernandoRios The last round bracket is missing in your code and the variable names are in the wrong case

reg Y X1 X2 [pw=normalden(Z,1.2375,53)]

Aryan Bose I still think that you should show us exactly what the reviewer has recommended to you. Demanding a simultaneous quintile regression based on the quintiles of an independent variable seems strange for me. I have never heard of something like this before. Therefore, I suspect that the reviewer might want to see something different than you believe and currently present here.

Dear FernandoRios and Sven-Kristjan,

Thank you for your suggestion. In one of my paper, I have shown that the positive association between firm performance and disclosure is moderated by the foreign operations. I have measured foreign operations as a binary variable based on the amount of foreign sales (if a firm makes foreign sales, then 1 otherwise 0). The regression equation I estimated:

reg Y X1*M X2 X3 X4 X5 X6 X7 X8 X9

M is a moderator (dummy variable) which is computed based on Z (Amount of Foreign sales).

The reviewer’s comment is:

"Furthermore, right now the authors measure foreign operations by the quantum of foreign sales. I would also appreciate seeing some supplementary analyses in which the relationship the authors examine alters or changes depending on the amount of foreign sales occurring. A simultaneous quintile model would work in this instance and provide very illuminating results."

Aryan

[Sven-Kristjan Bormann]

I would interpret the reviewer's comment as hinting at an analysis of subsamples where the subsamples are defined by the quintiles of the foreign sales, as already discussed in this thread.
I would be also irritated by asking for a "simultaneous quintile model". Maybe in the mind of the reviewer estimating the same model over differenent quintiles of an independent variable and presenting the results together counts as a "simultaneous quintile model".
Another alternative might be to use a categorical variable instead of dummy variable where the categorical variable takes the value 1 to 5 for the quintiles. You could play around with both approaches and see if they provide some useful insights.
Otherwise, I would just admit in your answer to the reviewer that you were a bit confused by the demand but that you tried to fullfil the demand according to your best understanding with these different approaches.

[Aryan Bose]

Many thanks Sven-Kristjan for your great suggestion and help.