Partially Adaptive Quantile Estimation

• Activity Number: 419
• Type: Contributed
• Date/Time: Tuesday, August 11, 2015 : 2:00 PM to 3:50 PM
• Sponsor: Business and Economic Statistics Section
• Author(s): James McDonald* and David J. Mauler
• Companies: Brigham Young University and Brigham Young University
• Keywords: Semiparametric ; Skewed Laplace ; Skewness ; Heteroskedasticity ; Monte Carlo

Abstract

This paper contrasts two approaches to estimating quantile regression models: traditional semiparametric methods and partially adaptive estimators using flexible probability density functions (pdfs). While more general pdfs could have been used, the skewed Laplace was selected for pedagogical purposes. Monte Carlo simulations are used to compare the behavior of the semiparametric and partially adaptive quantile estimators in the presence of possibly skewed and heteroskedastic data. Both approaches accommodate skewness and heteroskedasticity which are consistent with linear quantiles; however, the partially adaptive estimator considered allows for nonlinear quantiles and also provides simple tests for symmetry and heteroskedasticity. An application of the methods is also considered.