I am a beginner in quantile regression, quantreg package in R.I found that it is a good method to analyze data with outliers or non-normally disturbing data, but can´t find anything about homoscedasticity and autocorrelation assumptions. Research papers using quantile regression don´t mention it.

Can somebody help me with the assumptions that the model in quantile regression must fulfill? Are homoscedasticity and no autocorrelation important in this type of regression? I can not find any relevant literature about it. I know that heteroscedasticity and autocorrelation lead to the biased coefficient in the linear regression model and therefore I usually used variance-covariance matrix with appropriate arguments.

But how can I interpret the results of quantile regression, if there is a problem with autocorrelation, heteroscedasticity, or both of them?