CMStatistics 2020: Start Registration
View Submission - CFE
Title: Forecasting value-at-risk and expected shortfall: A Bayesian approach Authors:  J Miguel Marin - University Carlos III (Spain) [presenting]
Helena Veiga - UNED (Spain)
Abstract: The main aim is to investigate whether the asymmetric response of the volatility plays an important role in forecasting the short term horizon VaR and ES and whether the choice of the model determines these forecasts. There is some evidence in the literature reporting that volatility asymmetries included in GARCH type models improve VaR and ES for short horizons. We extend the literature by modeling five international stock market returns according to six volatility models: three belong to the GARCH family while the other three models are in the SV family. All models estimations, VaR and ES forecasts are obtained with a Markov chain Monte Carlo (MCMC) procedure implemented in the R package Nimble and package doParallel that allows for parallel computing. This is a great advantage given that MCMC methods are known to be time-consuming.