A cross-country comparison of Expected Shortfall estimation models

This dissertation aims to find a better-performed model in estimating risk measures for certain countries. The risk measures are estimated under five distributional assumptions (normal, Student-t, skewed Student-t, historical distribution, and generalized pareto) for five financial markets (Nasdaq,...

Full description

Bibliographic Details
Main Author: Du, Cong
Format: Dissertation (University of Nottingham only)
Language:English
Published: 2017
Online Access:https://eprints.nottingham.ac.uk/46082/
_version_ 1848797253897027584
author Du, Cong
author_facet Du, Cong
author_sort Du, Cong
building Nottingham Research Data Repository
collection Online Access
description This dissertation aims to find a better-performed model in estimating risk measures for certain countries. The risk measures are estimated under five distributional assumptions (normal, Student-t, skewed Student-t, historical distribution, and generalized pareto) for five financial markets (Nasdaq, FTSE100, SSEC, BVSP, and Nifty50), three estimation windows (250, 500, 1000), and two significance levels (0.05 and 0.01). A two-stage ES backtest and a direct ES test are conducted. The violation ratio comparisons and the backtesting results indicate that student’s t distribution and normal distribution have overall the best and the worst performance. Unconditional models also have its deficiencies in modeling financial markets.
first_indexed 2025-11-14T20:00:57Z
format Dissertation (University of Nottingham only)
id nottingham-46082
institution University of Nottingham Malaysia Campus
institution_category Local University
language English
last_indexed 2025-11-14T20:00:57Z
publishDate 2017
recordtype eprints
repository_type Digital Repository
spelling nottingham-460822018-04-17T15:07:57Z https://eprints.nottingham.ac.uk/46082/ A cross-country comparison of Expected Shortfall estimation models Du, Cong This dissertation aims to find a better-performed model in estimating risk measures for certain countries. The risk measures are estimated under five distributional assumptions (normal, Student-t, skewed Student-t, historical distribution, and generalized pareto) for five financial markets (Nasdaq, FTSE100, SSEC, BVSP, and Nifty50), three estimation windows (250, 500, 1000), and two significance levels (0.05 and 0.01). A two-stage ES backtest and a direct ES test are conducted. The violation ratio comparisons and the backtesting results indicate that student’s t distribution and normal distribution have overall the best and the worst performance. Unconditional models also have its deficiencies in modeling financial markets. 2017-09-13 Dissertation (University of Nottingham only) NonPeerReviewed application/pdf en https://eprints.nottingham.ac.uk/46082/1/A%20cross-country%20comparison%20of%20es%20final%20.pdf Du, Cong (2017) A cross-country comparison of Expected Shortfall estimation models. [Dissertation (University of Nottingham only)]
spellingShingle Du, Cong
A cross-country comparison of Expected Shortfall estimation models
title A cross-country comparison of Expected Shortfall estimation models
title_full A cross-country comparison of Expected Shortfall estimation models
title_fullStr A cross-country comparison of Expected Shortfall estimation models
title_full_unstemmed A cross-country comparison of Expected Shortfall estimation models
title_short A cross-country comparison of Expected Shortfall estimation models
title_sort cross-country comparison of expected shortfall estimation models
url https://eprints.nottingham.ac.uk/46082/