Tail Risk Analysis Using realized measure and Dynamic Asymmetric Laplace Models in Tehran Stock Exchange

The main objective of this study is to estimate and evaluate the performance of the dynamic realized conditioned autoregressive value at risk model (Realized-ES-CAViaR-Add-RV-SAV) in forecasting tail risk measures (VaR and ES). In this regard, daily as well as intraday (hourly) data of Tehran Stock Exchange Index in the period of 24/6/2014 – 2/2/2021are used. The results of the model are compared to the results of ES-CAViaR-SAV and ES-CAViaR-AS models to investigate the effect of incorporating the realized component to the model. Using backtesting tools such as Bin, POF, TUFF, CC, CCI, VRate tests, Lopez loss function (LL) (in VaR part) and McNeil and Frey test and ranking according to MCS method in The ES part, the efficiency of the models are examined. The results of this study indicate the efficiency of all three models in forecasting the tail risk measures. In addition, the results show that the use of realized criteria increases the tail risk forecasting efficiency.