Modellrisiko bei Value at Risk-Schätzungen: ein Vergleich zwischen Normalverteilung und hyperbolischer Verteilung

This article shows on an empirical data basis that return distributions of stocks of the swiss stock market can be successfully described through the hyperbolic distribution. Comparing this distribution with the normal distribution one can achieve a more accurate modelling. In addition it is shown, that the use of a hyperbolic distribution for the measurement of market risk with the Value at Risk concept will lead to a reduction in model risk.