ID: 55858
Authors:
Jaime Enrique Lincovil, Chang Chiann.
Source:
Revista Brasileira de Finanças, v. 17, n. 4, p. 56-76, October-December, 2019. 21 page(s).
Keyword:
Backtesting , Conditional coverage , Empirical power , Expected shortfall , Value–at–risk
Document type: Article (Portuguese)
Show Abstract
The assessment of risk measures forecast, as the value–at–risk (VaR) and the expected shortfall (ES), is a relevant process for financial institutions. The backtestings were introduced to evaluate the efficiency of their forecasts. In this work, we compare the empirical power of new classes of backtestings, for VaR and ES, presented in the statistical literature. Further, we employ these procedures to evaluate the efficiency of the forecasts generated by both the Historical Simulation method and two methods based on the Generalized Pareto Distribution. In the order to evaluate the VaR forecast, the empirical power of the Geometric–VaR class of backtesting was, in general, higher than the others in the simulated scenarios. This supports the advantages of the employment of durations and covariates in the test procedures. On the other hand, to evaluate the ES forecast, the backtestings based on the conditional distribution of the returns to the VaR showed a good performance. Additionally, we show that the method based on the generalized Pareto distribution using durations and covariates results in an optimal performance in the forecast of VaR and ES according to the backtestings.
