Skip to main content

LUP Student Papers

LUND UNIVERSITY LIBRARIES

Backtesting Expected Shortfall A comparative empirical evaluation of different backtests

Johansson, Jesper LU and Fredriksson, Viktor LU (2020) NEKN01 20201
Department of Economics
Abstract
This paper empirically evaluates whether different backtests for Expected Shortfall (ES) produce similar results. In 2016, the Basel Committee on Banking Supervision proposed a shift from Value-at-Risk (VaR) to ES as the industry standard when calculating capital requirements for banks. However, ES has been found difficult to backtest. Since backtesting results form the basis for determining the capital requirements of banks it is important to elucidate whether the backtests produce similar results. We answer this question by performing six different daily backtests on the S&P 500 index for the period 1965-2020 and measuring correlations between the different backtests. We found a substantial divergence across different backtests. We also... (More)
This paper empirically evaluates whether different backtests for Expected Shortfall (ES) produce similar results. In 2016, the Basel Committee on Banking Supervision proposed a shift from Value-at-Risk (VaR) to ES as the industry standard when calculating capital requirements for banks. However, ES has been found difficult to backtest. Since backtesting results form the basis for determining the capital requirements of banks it is important to elucidate whether the backtests produce similar results. We answer this question by performing six different daily backtests on the S&P 500 index for the period 1965-2020 and measuring correlations between the different backtests. We found a substantial divergence across different backtests. We also found that the correlations remain stable or increase during the global financial crisis. In the light of these results we recommend practitioners to diversify between multiple backtests. (Less)
Please use this url to cite or link to this publication:
author
Johansson, Jesper LU and Fredriksson, Viktor LU
supervisor
organization
course
NEKN01 20201
year
type
H1 - Master's Degree (One Year)
subject
keywords
Expected Shortfall, Backtests, Value-at-Risk, Empirical, Risk
language
English
id
9024227
date added to LUP
2020-08-29 10:33:32
date last changed
2020-08-29 10:33:32
@misc{9024227,
  abstract     = {{This paper empirically evaluates whether different backtests for Expected Shortfall (ES) produce similar results. In 2016, the Basel Committee on Banking Supervision proposed a shift from Value-at-Risk (VaR) to ES as the industry standard when calculating capital requirements for banks. However, ES has been found difficult to backtest. Since backtesting results form the basis for determining the capital requirements of banks it is important to elucidate whether the backtests produce similar results. We answer this question by performing six different daily backtests on the S&P 500 index for the period 1965-2020 and measuring correlations between the different backtests. We found a substantial divergence across different backtests. We also found that the correlations remain stable or increase during the global financial crisis. In the light of these results we recommend practitioners to diversify between multiple backtests.}},
  author       = {{Johansson, Jesper and Fredriksson, Viktor}},
  language     = {{eng}},
  note         = {{Student Paper}},
  title        = {{Backtesting Expected Shortfall A comparative empirical evaluation of different backtests}},
  year         = {{2020}},
}