Skip to main content

LUP Student Papers

LUND UNIVERSITY LIBRARIES

Backtesting The Parametric & Non-Parametric Estimates Expected Shortfall

Eiken, Johan LU and Winchester, Oliver LU (2022) NEKN01 20221
Department of Economics
Abstract
This paper will examine the forecast ability of different parametric and non-parametric estimation approaches used for Expected Shortfall (ES). As the financial institutions start to shift towards using ES as a risk measure instead of Value at Risk (VaR), the importance of being able to forecast ES increases. The Paper considers Basic Historical Simulation (BHS), Age-Weighted Historical Simulation (AWHS), Volatility-Weighted Historical Simulation (VWHS) and a parametric model based on a Normal distribution as well as t-distribution. These approaches were implemented over OMXS30 and OMXSPI from 1993-2022, as well as Bitcoin from 2011-2022. The results suggest that the TWHH and VWHS are the most reliable approaches in the context of... (More)
This paper will examine the forecast ability of different parametric and non-parametric estimation approaches used for Expected Shortfall (ES). As the financial institutions start to shift towards using ES as a risk measure instead of Value at Risk (VaR), the importance of being able to forecast ES increases. The Paper considers Basic Historical Simulation (BHS), Age-Weighted Historical Simulation (AWHS), Volatility-Weighted Historical Simulation (VWHS) and a parametric model based on a Normal distribution as well as t-distribution. These approaches were implemented over OMXS30 and OMXSPI from 1993-2022, as well as Bitcoin from 2011-2022. The results suggest that the TWHH and VWHS are the most reliable approaches in the context of estimating ES. (Less)
Please use this url to cite or link to this publication:
author
Eiken, Johan LU and Winchester, Oliver LU
supervisor
organization
course
NEKN01 20221
year
type
H1 - Master's Degree (One Year)
subject
keywords
Expected Shortfall, Value-at-Risk, Backtest, Historical Simulation
language
English
id
9084303
date added to LUP
2022-10-10 09:21:38
date last changed
2022-10-10 09:21:38
@misc{9084303,
  abstract     = {{This paper will examine the forecast ability of different parametric and non-parametric estimation approaches used for Expected Shortfall (ES). As the financial institutions start to shift towards using ES as a risk measure instead of Value at Risk (VaR), the importance of being able to forecast ES increases. The Paper considers Basic Historical Simulation (BHS), Age-Weighted Historical Simulation (AWHS), Volatility-Weighted Historical Simulation (VWHS) and a parametric model based on a Normal distribution as well as t-distribution. These approaches were implemented over OMXS30 and OMXSPI from 1993-2022, as well as Bitcoin from 2011-2022. The results suggest that the TWHH and VWHS are the most reliable approaches in the context of estimating ES.}},
  author       = {{Eiken, Johan and Winchester, Oliver}},
  language     = {{eng}},
  note         = {{Student Paper}},
  title        = {{Backtesting The Parametric & Non-Parametric Estimates Expected Shortfall}},
  year         = {{2022}},
}