Biased Mean Quadrangle and Applications

TL;DR

This paper introduces biased mean regression, a method for estimating biased means in various applications, linking it to quantile regression and CVaR optimization, with computational advantages via linear programming.

Contribution

It develops the biased mean quadrangle framework, establishing theoretical properties and connections to existing paradigms, and demonstrates practical benefits through numerical experiments.

Findings

Biased mean regression is equivalent to quantile regression for certain parameters.

Minimizing superexpectation error reduces to linear programming.

Numerical experiments confirm theoretical properties and practical utility.

Abstract

This paper introduces \emph{biased mean regression}, estimating the \emph{biased mean}, i.e., $\mathbb{E}[Y] + x$, where $x \in \mathbb{R}$. The approach addresses a fundamental statistical problem that covers numerous applications. For instance, it can be used to estimate factors driving portfolio loss exceeding the expected loss by a specified amount (e.g., $ x=\$10 billion$) or to estimate factors impacting a specific excess release of radiation in the environment, where nuclear safety regulations specify different severity levels. The estimation is performed by minimizing the so-called \emph{superexpectation error}. We establish two equivalence results that connect the method to popular paradigms: (i) biased mean regression is equivalent to quantile regression for an appropriate parameterization and is equivalent to ordinary least squares when $x=0$; (ii) in portfolio optimization, minimizing \emph{superexpectation risk}, associated with the superexpectation error, is equivalent to CVaR optimization. The approach is computationally attractive, as minimizing the superexpectation error reduces to linear programming (LP), thereby offering algorithmic and modeling advantages. It is also a good alternative to ordinary least squares (OLS) regression. The approach is based on the \emph{Risk Quadrangle} (RQ) framework, which links four stochastic functionals -- error, regret, risk, and deviation -- through a statistic. For the biased mean quadrangle, the statistic is the biased mean. We study properties of the new quadrangle, such as \emph{subregularity}, and establish its relationship to the quantile quadrangle. Numerical experiments confirm the theoretical statements and illustrate the practical implications.