Resumen de Role of choice of threshold on the estimation of market risk under the pot method (EVT)

Sonia Benito Muela, Carmen López Martín, Mª Ángeles Navarro

• The conditional extreme value theory has been proven to be one of the most successful in estimating market risk. The implementation of this method in the framework of the Peaks Over a Threshold (POT) model requires one to choose a threshold for fitting the generalized Pareto distribution (GPD). In this paper, we investigate whether the selection of the threshold is important for the quantification of market risk. For measuring risk, we use the value at risk (VaR) measure. The study has been done for the S&P 500 index. When we analyse the validity of VaR estimates, the results show than all the thresholds considered provide correct estimations of VaR, with the exception of certain thresholds corresponding to the 99th percentile. The results obtained show that the quantification of the market risk through the VaR does not depend on the threshold selected.