Advanced

A comparative study of VaR models

Recht, Sebastian LU and Wang, Kaifeng LU (2012) NEKN02 20121
Department of Economics
Abstract (Swedish)
In this paper, we investigate the predictive performance of eight common Value at Risk (VaR) estimation methods with the daily data of OMX Stockholm 30 and FTSE 100 indexes. These methods include non-parametric, parametric as well as Extreme Value Theory approaches whose daily and 10-day estimates are then evaluated using the Christoffersen frequency test. The reason why we choose these two stock indexes is because that they respectively represent markets with small and big capitalizations, and the main goal of our study is to search for empirical evidences on whether or not the market size could have some important implications on the Value at Risk estimation process, such as if a universal approach could be applied to both types of... (More)
In this paper, we investigate the predictive performance of eight common Value at Risk (VaR) estimation methods with the daily data of OMX Stockholm 30 and FTSE 100 indexes. These methods include non-parametric, parametric as well as Extreme Value Theory approaches whose daily and 10-day estimates are then evaluated using the Christoffersen frequency test. The reason why we choose these two stock indexes is because that they respectively represent markets with small and big capitalizations, and the main goal of our study is to search for empirical evidences on whether or not the market size could have some important implications on the Value at Risk estimation process, such as if a universal approach could be applied to both types of markets. (Less)
Please use this url to cite or link to this publication:
author
Recht, Sebastian LU and Wang, Kaifeng LU
supervisor
organization
course
NEKN02 20121
year
type
H1 - Master's Degree (One Year)
subject
keywords
Christoffersen frequency test, Extreme value theory, Market capitalization, Historical simulation, Value at Risk
language
English
id
2621156
date added to LUP
2012-06-08 14:30:59
date last changed
2012-06-08 14:30:59
@misc{2621156,
  abstract     = {In this paper, we investigate the predictive performance of eight common Value at Risk (VaR) estimation methods with the daily data of OMX Stockholm 30 and FTSE 100 indexes. These methods include non-parametric, parametric as well as Extreme Value Theory approaches whose daily and 10-day estimates are then evaluated using the Christoffersen frequency test. The reason why we choose these two stock indexes is because that they respectively represent markets with small and big capitalizations, and the main goal of our study is to search for empirical evidences on whether or not the market size could have some important implications on the Value at Risk estimation process, such as if a universal approach could be applied to both types of markets.},
  author       = {Recht, Sebastian and Wang, Kaifeng},
  keyword      = {Christoffersen frequency test,Extreme value theory,Market capitalization,Historical simulation,Value at Risk},
  language     = {eng},
  note         = {Student Paper},
  title        = {A comparative study of VaR models},
  year         = {2012},
}