How should I choose the quantiles in quantile regression ?

[Marcos Gonzalez]

I am running a quantile regression where my dependent variable is the average debt maturity and I would like to know if there is a rule to select the quantiles of the regression. I am using quantiles 0.10 0.25 0.5 0.75 and 0.9 but I do not know if it is correct. Is there any paper where this is discussed?

Thanks in advance

[Nick Cox]

I can't help with the literature on average debt maturity. In general the answer should depend entirely on your substantive goals.

[Marcos Gonzalez]

Thank you for your answer.

So, I should check on the literature about this topic and try to find some references that have done a similar analysis and use the same quantiles. The thing is that I have already done this and I have not found a similar analysis for this topic. The objective of my paper is to analyze the main determinants of debt maturity for a set of international companies with different sizes and characteristics. So, I have a long distribution of values for the variables and I think that quantile regression better fits with my distribution of data and I would like to know the relationship between my dependent and explanatory variables for the whole distribution to see if the relationship changes for different values of the dependent variable, that is the reason to take these quantiles that summarize, more or less, the whole distribution.

[Nick Cox]

The test lies in your results. If you have results that make sense and are well supported, they will be interesting and useful to others.