Value-at-Risk analysis of stock returns: Historical simulation, varinace techniques or tail index estimation ?

In this paper various Value-at-Risk techniques are applied tot the Dutch stock market index AEX and to the Dow Jones Industrial Average. the main conclusions are: (1) Changing volatility over time is the most important characteristic of stock returns when modelling value-at-risk; (2) For high confidence levels, the fat tails of the distribution can best be modeled by means of the t-distribution, tail index estimators perform much worse; (3) The penalty scheme used to determine the capital requirement is not discriminating enough to give banks proper incentives.