A Single Risk Approach to the Semiparametric Competing Risks Model With Parametric Archimedean Risk Dependence

Simon M. S.  Lo; Ralf Wilke

doi:10.1016/j.jmva.2023.105276

A Single Risk Approach to the Semiparametric Competing Risks Model With Parametric Archimedean Risk Dependence

Simon M. S. Lo, Ralf Wilke^*

^*Corresponding author for this work

Department of Economics

United Arab Emirates University

Research output: Contribution to journal › Journal article › Research › peer-review

47 Downloads (Pure)

Abstract

This paper considers a dependent competing risks model with the distribution of one risk being a semiparametric proportional hazards model, whereas the model for the other risks and the degree of risk dependence of an Archimedean copula are unknown. Identifiability is shown when there is at least one covariate with at least two values. Estimation is done by means of a -consistent semiparametric two-step procedure. Applicability and attractive finite sample performance are demonstrated with the help of simulations. An application to unemployment duration confirms the importance of estimating rather than assuming risk dependence.

Original language	English
Article number	105276
Journal	Journal of Multivariate Analysis
Volume	201
Number of pages	13
ISSN	0047-259X
DOIs	https://doi.org/10.1016/j.jmva.2023.105276
Publication status	Published - May 2024

Bibliographical note

Published online: 24 November 2023.

Keywords

Archimedean copula
Depending censoring
Identifiability

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1016/j.jmva.2023.105276Licence: CC BY

simon_m_s_lo_et_al_a_single_risk_approach_publishersversionProof, 613 KBLicence: CC BY

Handle.net

Persistent link

Cite this

@article{4fb4418649ee49ddb752c9effc71d997,

title = "A Single Risk Approach to the Semiparametric Competing Risks Model With Parametric Archimedean Risk Dependence",

abstract = "This paper considers a dependent competing risks model with the distribution of one risk being a semiparametric proportional hazards model, whereas the model for the other risks and the degree of risk dependence of an Archimedean copula are unknown. Identifiability is shown when there is at least one covariate with at least two values. Estimation is done by means of a -consistent semiparametric two-step procedure. Applicability and attractive finite sample performance are demonstrated with the help of simulations. An application to unemployment duration confirms the importance of estimating rather than assuming risk dependence.",

keywords = "Archimedean copula, Depending censoring, Identifiability, Archimedean copula, Depending censoring, Identifiability",

author = "Lo, {Simon M. S.} and Ralf Wilke",

note = "Published online: 24 November 2023.",

year = "2024",

month = may,

doi = "10.1016/j.jmva.2023.105276",

language = "English",

volume = "201",

journal = "Journal of Multivariate Analysis",

issn = "0047-259X",

publisher = "Elsevier",

}

TY - JOUR

T1 - A Single Risk Approach to the Semiparametric Competing Risks Model With Parametric Archimedean Risk Dependence

AU - Lo, Simon M. S.

AU - Wilke, Ralf

N1 - Published online: 24 November 2023.

PY - 2024/5

Y1 - 2024/5

N2 - This paper considers a dependent competing risks model with the distribution of one risk being a semiparametric proportional hazards model, whereas the model for the other risks and the degree of risk dependence of an Archimedean copula are unknown. Identifiability is shown when there is at least one covariate with at least two values. Estimation is done by means of a -consistent semiparametric two-step procedure. Applicability and attractive finite sample performance are demonstrated with the help of simulations. An application to unemployment duration confirms the importance of estimating rather than assuming risk dependence.

AB - This paper considers a dependent competing risks model with the distribution of one risk being a semiparametric proportional hazards model, whereas the model for the other risks and the degree of risk dependence of an Archimedean copula are unknown. Identifiability is shown when there is at least one covariate with at least two values. Estimation is done by means of a -consistent semiparametric two-step procedure. Applicability and attractive finite sample performance are demonstrated with the help of simulations. An application to unemployment duration confirms the importance of estimating rather than assuming risk dependence.

KW - Archimedean copula

KW - Depending censoring

KW - Identifiability

KW - Archimedean copula

KW - Depending censoring

KW - Identifiability

U2 - 10.1016/j.jmva.2023.105276

DO - 10.1016/j.jmva.2023.105276

M3 - Journal article

SN - 0047-259X

VL - 201

JO - Journal of Multivariate Analysis

JF - Journal of Multivariate Analysis

M1 - 105276

ER -