Quantile-on-Quantile Regression

[Odongo Kodongo]

Hello. Does anyone here have the code for running Quantile-on-Quantile regression (Sim and Zhou, 2015) in Stata? Thank you.

Sim, N., and Zhou, H., 2015. Oil prices, US stock return, and the dependence between their quantiles. Journal of Banking and Finance 55, pp. 1 - 8. http://dx.doi.org/10.1016/j.jbankfin.2015.01.013

[FernandoRios]

Hi Odongo
There is no command for this Quantile on quantile regression. However, my reading of the paper suggests the regression is simply a modification of the semiparametric varying coefficient model, but applied to quantile regressions, rather than linear regression.
The paper provides no recommendation on the estimation of bandwidths, thus, I think you can estimate it manually simply defining the appropriate kernel weights, and run the qregression with those weights (and a transformed variable)
HTH
F

Edited:
You could also estimate the same using a more simple approach : splines
basically
q(y|t,X,z)=a(t)+b(t)*x+f(z,t)
where f(z,t) is just a flexible function of Z.
Marginal effects can then be estimated using margins or "f_able" (ssc)

HTH