Disentangling distortion risk measures and the Expected Shortfall

Research output: Working paper/Preprint › Working paper/Discussion paper

Authors

  • Felix-Benedikt Liebrich
  • Massimiliano Amarante

Research Organisations

External Research Organisations

  • University of Montreal
View graph of relations

Details

Original languageEnglish
Number of pages27
Publication statusE-pub ahead of print - 2022

Abstract

Distortion risk measures are risk measures that are law invariant and comonotonic addi- tive. We characterize notable sub-classes of this wide range of functionals, starting with the property of prudence recently introduced by Wang & Zitikis. Moreover, we develop a new view of coherent distortion risk measures and show that they are captured by a single probability charge. By linking our insights into these two properties, we obtain new characterizations of the Expected Shortfall and implications associated with its utilization as standard measure of market risk. Along the route, we obtain some ancillary results of independent interest. These concern: (i) the anticore of a general submodular distortion; (ii) a full characterization of spectral risk measures on integrable random variables; and (iii) a novel proof of the automatic Fatou property of convex, law invariant risk measures. Finally, we fully close the remaining gap to the Wang-Zitikis axiomatization of the Expected Shortfall and carefully disentangle the interplay of the involved axioms within the large class of distortion risk measures.

Cite this

Disentangling distortion risk measures and the Expected Shortfall. / Liebrich, Felix-Benedikt; Amarante, Massimiliano.
2022.

Research output: Working paper/Preprint › Working paper/Discussion paper

Liebrich FB, Amarante M. Disentangling distortion risk measures and the Expected Shortfall. 2022. Epub 2022.
Liebrich, Felix-Benedikt ; Amarante, Massimiliano. / Disentangling distortion risk measures and the Expected Shortfall. 2022.
Download
@techreport{34230198b42d46b7846b7604b0c56387,
title = "Disentangling distortion risk measures and the Expected Shortfall",
abstract = "Distortion risk measures are risk measures that are law invariant and comonotonic addi- tive. We characterize notable sub-classes of this wide range of functionals, starting with the property of prudence recently introduced by Wang & Zitikis. Moreover, we develop a new view of coherent distortion risk measures and show that they are captured by a single probability charge. By linking our insights into these two properties, we obtain new characterizations of the Expected Shortfall and implications associated with its utilization as standard measure of market risk. Along the route, we obtain some ancillary results of independent interest. These concern: (i) the anticore of a general submodular distortion; (ii) a full characterization of spectral risk measures on integrable random variables; and (iii) a novel proof of the automatic Fatou property of convex, law invariant risk measures. Finally, we fully close the remaining gap to the Wang-Zitikis axiomatization of the Expected Shortfall and carefully disentangle the interplay of the involved axioms within the large class of distortion risk measures.",
author = "Felix-Benedikt Liebrich and Massimiliano Amarante",
year = "2022",
language = "English",
type = "WorkingPaper",

}

Download

TY - UNPB

T1 - Disentangling distortion risk measures and the Expected Shortfall

AU - Liebrich, Felix-Benedikt

AU - Amarante, Massimiliano

PY - 2022

Y1 - 2022

N2 - Distortion risk measures are risk measures that are law invariant and comonotonic addi- tive. We characterize notable sub-classes of this wide range of functionals, starting with the property of prudence recently introduced by Wang & Zitikis. Moreover, we develop a new view of coherent distortion risk measures and show that they are captured by a single probability charge. By linking our insights into these two properties, we obtain new characterizations of the Expected Shortfall and implications associated with its utilization as standard measure of market risk. Along the route, we obtain some ancillary results of independent interest. These concern: (i) the anticore of a general submodular distortion; (ii) a full characterization of spectral risk measures on integrable random variables; and (iii) a novel proof of the automatic Fatou property of convex, law invariant risk measures. Finally, we fully close the remaining gap to the Wang-Zitikis axiomatization of the Expected Shortfall and carefully disentangle the interplay of the involved axioms within the large class of distortion risk measures.

AB - Distortion risk measures are risk measures that are law invariant and comonotonic addi- tive. We characterize notable sub-classes of this wide range of functionals, starting with the property of prudence recently introduced by Wang & Zitikis. Moreover, we develop a new view of coherent distortion risk measures and show that they are captured by a single probability charge. By linking our insights into these two properties, we obtain new characterizations of the Expected Shortfall and implications associated with its utilization as standard measure of market risk. Along the route, we obtain some ancillary results of independent interest. These concern: (i) the anticore of a general submodular distortion; (ii) a full characterization of spectral risk measures on integrable random variables; and (iii) a novel proof of the automatic Fatou property of convex, law invariant risk measures. Finally, we fully close the remaining gap to the Wang-Zitikis axiomatization of the Expected Shortfall and carefully disentangle the interplay of the involved axioms within the large class of distortion risk measures.

M3 - Working paper/Discussion paper

BT - Disentangling distortion risk measures and the Expected Shortfall

ER -