A Semi-Parametric Method to Estimate VaR with EVT and ES

Value-at-Risk (VaR) is probably the most widely used risk measure in financial institutions, which is a quantile of the loss distribution in nature. In this article, a semi-parametric method which is based on the Extreme Value Theory(EVT) and the Historical Simulation(HS) is proposed to estimate VaR. We use EVT to estimate the tails and HS to simulate the interior. An empirical study using the daily returns of S&P 500 index is also presented. It's shown that that our approach is an adequate risk measure.