Value at Risk Forecasting Based on Quantile Regression for GARCH Models

Value-at-Risk(VaR) is an important part of risk management in the financial industry. This paper presents a VaR forecasting for financial time series based on the quantile regression for GARCH models recently developed by Lee and Noh (2009). The proposed VaR forecasting features the direct conditional quantile estimation for GARCH models that is well connected with the model parameters. Empirical performance is measured by several backtesting procedures, and is reported in comparison with existing methods using sample quantiles.