Testing applicability of value at risk models in stocks markets

Abstract

This paper evaluates the forecasting performance of Value at Risk (VaR) method based on two wide spread approaches, historical simulation and Risk Metrics, before and after the sub-prime crisis in the context of developed and emerging capital markets. We present results on both VaR 1% and VaR 5% on a one-day horizon for Belex 15 and SAX. For comparative purposes, the paper also focuses on the DJIA and the STOXX Eastern Europe Total Market Index, an index representative of emerging European stock markets. In order to validate accuracy of VaR results we employ different back test techniques. Results indicate that the relative performance of VaR as a measure of market risk significantly underestimates the true level of market risk in Serbian stock market, in contrast to Slovak, where standard VaR approaches accurately capture market risk exposure. Results also provide evidence that the characteristic of stock markets and their asset returns in combination with the desired confidence level and risk horizon determine how well a certain approach performs on a certain stock market.