Financial crises have become a principal concern to lead the development of new market risk indicators. The central limit theorem, which describes the average asymptotic behavior of a random process, does not characterize rare or extreme events, like subprime mortgage crises do. Value at Risk (VaR) is defined by risk exposure at a given probability level at a specified time horizon. Computing extreme value theory (EVT), focusing on the tails of the sample distribution, is an excellent approach for its use in managing risks. This research paper presents an application of extreme value theory to compute to Value-at-Risk of a market position in order to provide a consistent risk measurment.
KEY-WORDS : QUANTITATIVE FINANCE ; MATHEMATICAL MODELING ; PROBABILITY AND STATISTICS ; RISK MANAGEMENT ; EXTREME VALUE THEORY ; VALUE-AT-RISK (VAR)