Value-at-Risk (VaR) is a statistical metric that measures the risk of financial investments. More precisely, VaR measures the maximum amount one could lose over a specific time horizon, given a pre-defined confidence level. This is especially important for risk assessments and mitigations. In this talk, a forecasting method of the VaR of a portfolio will be discussed. As stocks tend to share some co-movement and dependence, the modeling is done using a high-dimensional vine copula model. More specifically, each stock of the portfolio is modeled using an ARMA-GARCH model while the dependence among the components is captured using a regular vine copula model, then the VaR is computed as an empirical quantile function. The simulation of a portfolio with 30 stocks taken from the 2008 financial crises period will be discussed.      «

Value-at-Risk (VaR) is a statistical metric that measures the risk of financial investments. More precisely, VaR measures the maximum amount one could lose over a specific time horizon, given a pre-defined confidence level. This is especially important for risk assessments and mitigations. In this talk, a forecasting method of the VaR of a portfolio will be discussed. As stocks tend to share some co-movement and dependence, the modeling is done using a high-dimensional vine copula model. More s...     »