Empirical Study of Value-at-Risk: application of different methodologies to Asian stock market indexes

Abstract

In my work I evaluate the performance of different VaR approaches applied on three stock market indexes. The estimation of these risk measures is compared and back-tested, in order to fulfill regulatory requirements from Basel Committee. The back-testing results highlight that the Historical Simulation, Filtered Historic Simulation and RiskMetrics underperform compared with the GARCH-N and GJR-T models; the first ones show the highest percentage of failure and remain in the red and yellow zone of the Basel II categorization tests.