Distributional vs. Quantile Regression

Given a scalar random variable Y and a random vector X defined on the same probability space, the conditional distribution of Y given X can be represented by either the conditional distribution function or the conditional quantile function. To these equivalent representations correspond two alternative approaches to estimation. One approach, distributional regression (DR), is based on direct estimation of the conditional distribution function; the other approach, quantile regression (QR), is instead based on direct estimation of the conditional quantile function. Indirect estimates of the conditional quantile function and the conditional distribution function may then be obtained by inverting the direct estimates obtained from either approach. Despite the growing attention to the DR approach, and the vast literature on the QR approach, the link between the two approaches has not been explored in detail. The aim of this paper is to fill-in this gap by providing a better understanding of the relative performance of the two approaches, both asymptotically and in finite samples, under the linear location model and certain types of heteroskedastic location-scale models.

[1]  A. Azzalini.  A note on the estimation of a distribution function and quantiles by a kernel method , 1981 .

[2]  Victor Chernozhukov,et al.  Quantile Regression Under Misspecification, with an Application to the U.S. Wage Structure , 2004 .

[3]  P. K. Bhattacharya,et al.  Kernel and Nearest-Neighbor Estimation of a Conditional Quantile , 1990 .

[4]  Pin T. Ng,et al.  Quantile smoothing splines , 1994 .

[5]  R. Serfling.  Approximation Theorems of Mathematical Statistics , 1980 .

[6]  Roger Koenker,et al.  Inference on the Quantile Regression Process , 2000 .

[7]  Roger Koenker,et al.  Inference for Quantile Regression , 2005 .

[8]  Pandu R. Tadikamalla,et al.  A Probability Distribution and its Uses in Fitting Data , 1979 .

[9]  Dominik Wied,et al.  Misspecification Testing in a Class of Conditional Distributional Models , 2013 .

[10]  Young K. Truong,et al.  Asymptotic Properties of Kernel Estimators Based on Local Medians , 1989 .

[11]  V. Chernozhukov,et al.  QUANTILE AND PROBABILITY CURVES WITHOUT CROSSING , 2007, 0704.3649.

[12]  E. Nadaraya,et al.  Some New Estimates for Distribution Functions , 1964 .

[13]  V. Chernozhukov,et al.  Inference on Counterfactual Distributions , 2009, 0904.0951.

[14]  F. Peracchi,et al.  The Conditional Distribution of Excess Returns: An Empirical Analysis , 1994 .

[15]  Quantile Regression under Misspecification , 2004 .

[16]  Franco Peracchi.  On estimating conditional quantiles and distribution functions , 2002 .

[17]  C. Rothe.  Partial Distributional Policy Effects , 2011, SSRN Electronic Journal.

[18]  Nonparametric regression M-quantiles , 1989 .

[19]  Jon A. Wellner,et al.  Weak Convergence and Empirical Processes: With Applications to Statistics , 1996 .

[20]  R. Koenker.  Quantile Regression: Fundamentals of Quantile Regression , 2005 .

[21]  Holger Dette,et al.  Non‐crossing non‐parametric estimates of quantile curves , 2008 .

[22]  R. Spady,et al.  AN EFFICIENT SEMIPARAMETRIC ESTIMATOR FOR BINARY RESPONSE MODELS , 1993 .

[23]  R. Koenker,et al.  Regression Quantiles , 2007 .

[24]  R. Koenker,et al.  Hierarchical Spline Models for Conditional Quantiles and the Demand for Electricity , 1990 .

[25]  M. Samanta,et al.  Non-parametric estimation of conditional quantiles , 1989 .

[26]  Probal Chaudhuri,et al.  Nonparametric Estimates of Regression Quantiles and Their Local Bahadur Representation , 1991 .