Assessing the Performance of Parametric, Non-Parametric and Semi-Parametric Value-at-Risk Models Applied to the Chinese Stock Market

PENG, BO (2006) Assessing the Performance of Parametric, Non-Parametric and Semi-Parametric Value-at-Risk Models Applied to the Chinese Stock Market. [Dissertation (University of Nottingham only)] (Unpublished)

[img] PDF - Registered users only - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
Download (775kB)

Abstract

In this paper, parametric, nonparametric, and semiparametric models are applied to a hypothetical portfolio consisting a single asset-Shanghai Stock Index 180, to assess their performance in the Chinese stock market. Some stylized facts and features of stock returns have been documented by many empirical studies, and which have been found to be important for VaR estimations. The main findings of this paper are: the fat tailness in stock return distribution is the most important stylized fact for increasing the accuracy of VaR estimations, while leverage effects is found to be most important for properly modelling the time-series behaviour of VaR estimations. Among all the tested models, based on whole performance, GARCH (1.1)-t (d) model is found to be the most appropriate model for measuring the risk of the Chinese stock market.

Item Type: Dissertation (University of Nottingham only)
Keywords: Value at risk, VaR, GARCH, Historical simulation
Depositing User: EP, Services
Date Deposited: 13 Apr 2007
Last Modified: 23 Oct 2016 12:14
URI: http://eprints.nottingham.ac.uk/id/eprint/20211

Actions (Archive Staff Only)

Edit View Edit View