Seminar by Wenxin Zhou, University of Illinois Chicago

Abstract

Expected Shortfall (ES), also known as superquantile or Conditional Value-at-Risk, is recognized as an important risk measure in risk management and decision making. In this talk, we consider a joint regression framework that simultaneously models the conditional quantile and ES of a response variable given a set of covariates. The state-of-the-art approach for this framework involves minimizing a joint loss function that is non-differentiable and non-convex. Motivated by the idea of using orthogonal scores to reduce sensitivity to nuisance parameters, we study a unified two-step framework for fitting joint quantile and ES regression models, applicable from linear to nonlinear settings. We also discuss a natural robust approach to make the estimator more resistant to heavy-tailed response distributions. A Python package, named quantes (https://pypi.org/project/quantes/), has been developed to implement ES regressions.

Bio

Wenxin Zhou joined the Department of Information and Decision Sciences at the University of Illinois at Chicago in 2023, after spending six years in the Department of Mathematics at UCSD. His research interests include high-dimensional inference, robust learning with heavy-tailed data, nonparametric statistics, and quantile and expected shortfall regression.