Value-at-risk. Measurement and evaluation methods for market risk

The prudential regulation of financial institutions requires the maintenance of minimum levels of capital as reserves against financial risks. Banks now have the option to use their own VaR risk-management model as the basis for required capital ratios. The main objective of this paper is to illustrate a VaR methodology that could be used by the banks in elaborating their internal models such as: standard GARCH, GJR and EGARCH models under three distributional assumptions (normal, GED and Student-t). VaR models are useful as they can be demonstrated to be reasonably accurate. To do this, we must check systematically the validity of the underlying valuation and risk models through comparison of predicted and actual loss levels, namely backtesting.