Welcome to Ruodu Wang's Homepage

Lhasa, China, 2019

Working Paper Series

Axiomatic theory of risk measures Risk sharing, allocation, and equilibria
Joint mixability Robust risk aggregation
Risk management with risk measures E-values: Theory and methods

Risk sharing, allocation, and equilibria

The series of working papers contains recent advances on risk sharing problems and equilibria. Objectives, preferences, and constraints in most papers are modelled by risk measures, with a special focus on the industry standards such as the Value-at-Risk (VaR), the Expected Shortfall (ES, also known as TVaR, CVaR and AVaR) and the Range-Value-at-Risk (RVaR, WP02). In particular, the framework of quantile-based risk sharing is initiated in WP02, which is further explored in WP04, WP06 and WP08. In addition to risk allocation and financial market equilibria, insurance and reinsurance arrangements are also common forms of risk sharing in an incomplete market (WP03, WP05).

The inf-convolution is a common mathematical tool to study Pareto and competitive equilibria. Technical challenges often arise from the non-convexity of most of the interesting objectives (such as VaR, RVaR and other distortion risk measures), whereas inf-convolution and risk sharing are extensively studied in the literature with convexity.

Most papers provide theoretical results with the hope to offer insights for regulation and risk management practice. A brief description of each working paper is provided below to explain its main results and the logical structure across papers.

WP10 An unexpected stochastic dominance: Pareto distributions, catastrophes, and risk exchange
(by Yuyu Chen, Paul Embrechts, Ruodu Wang)

We show the perhaps surprising inequality (Theorem 1) that the weighted average of iid ultra heavy-tailed (i.e., infinite mean) Pareto losses is larger than a standalone loss in the sense of first-order stochastic dominance. We discuss several implications of these results via an equilibrium analysis in a risk exchange market.

WP09 An axiomatic theory to anonymized risk sharing
(by Zhanyi Jiao, Yang Liu, Ruodu Wang)

We study an axiomatic framework for anonymized risk sharing. Four natural axioms, actuarial fairness, risk fairness, risk anonymity, and operational anonymity, are proposed and discussed. We establish in Theorem 1 that the four axioms characterizes the conditional mean risk sharing rule, revealing the unique and prominent role of this popular risk sharing rule among all others in relevant applications of anonymized risk sharing. A characterization of the conditional expectation in Theorem 2 may be of independent interest.

WP08 Inf-convolution, optimal allocations, and model uncertainty for tail risk measures
(by Fangda Liu, Tiantian Mao, Ruodu Wang, Linxiao Wei)

This paper is the fourth in a series of papers on quantile-based risk sharing. We extend the setting of WP02 to general tail risk measures (coined in WP04 of Axiomatic theory of risk measures) based on results from WP07, with a few technical results distinguishing the left-VaR and the right-VaR. Two types of model uncertainty are studied in the setting of risk sharing, and useful formulas are obtained on worst-case values of risk measures under uncertainty modeled by the Wasserstein metric. The journal version is to appear in Mathematics of Operations Research (2021).

WP07 Is the inf-convolution of law-invariant preferences law-invariant?
(by Peng Liu, Ruodu Wang, Linxiao Wei)

To the question reflected in the paper's title, an answer is generally negative even in an atomless probability space, but it becomes positive with minor assumptions. The paper contains many subtle examples. The journal version is published in Insurance: Mathematics and Economics (2020).

WP06 Characterizing optimal allocations in quantile-based risk sharing
(by Ruodu Wang, Yunran Wei)

This paper is the third in a series of papers on quantile-based risk sharing. Via several technical results, the paper offers results on existence, uniqueness, and characterization of Pareto-optimal allocations considered in WP02. In addition, it contains some auxiliary results on VaR and ES optimization. The journal version is published in Insurance: Mathematics and Economics (2020).

WP05 Competitive equilibria in a comonotone market
(by Tim Boonen, Fangda Liu, Ruodu Wang)

This paper is a study of competitive equilibria in a comonotone market and their relation to equilibria in a complete financial market. Many of our results are in sharp contrast to the ones on complete markets in terms of existence, uniqueness, and closed-form solutions of the equilibria, and monotonicity of the pricing kernel. The journal version is to appear in Economic Theory (2020).

WP04 Quantile-based risk sharing with heterogeneous beliefs
(by Paul Embrechts, Haiyan Liu, Tiantian Mao, Ruodu Wang)

This paper is the second in a series of papers on quantile-based risk sharing, by generalizing WP02 to the setting of heterogeneous beliefs. Pareto-optimal and equilibrium allocations for ES agents are obtained in explicit formulas (Theorems 1-3), whereas they are inexplicit for VaR agents. The journal version is published in Mathematical Programming (2020).

WP03 Pareto-optimal reinsurance arrangements under general model settings
(by Jun Cai, Haiyan Liu, Ruodu Wang)

In this paper is a treatment of Pareto optimality in the context of reinsurance arrangements. Main results are necessary and sufficient conditions for Pareto-optimality and its existence, which are based on classic results in game theory. The journal version is published in Insurance: Mathematics and Economics (2017).

WP02 Quantile-based risk sharing
(by Paul Embrechts, Haiyan Liu, Ruodu Wang)

This paper is the first in a series of papers on quantile-based risk sharing. The main focus is Pareto and Arrow-Debreu equilibria for VaR, ES and RVaR. Main results include a fundamental quantile inequality (Theorem 1) and explicit formulas for optimal allocation (Theorem 2) and for equilibria (Theorem 3). The journal version is published in Operations Research (2018).

WP01 Regulatory arbitrage of risk measures
(by Ruodu Wang)

Regulatory arbitrage is the amount of possible capital requirement reduction through splitting a financial risk into several fragments, and in this paper it is formulated via self-convolution of a risk measure. In particular, VaR has infinite regulatory arbitrage. The journal version is published in Quantitative Finance (2016).