Dominique Guégan, Raymond Brummelhuis
Let (rt)t be a general GARCH(1G 1)-type process. We give explicit integral formula's for the probability densities of rt+k conditional to given values of rt,stG, where the latter denotes the variance. As an application, we study the extreme value asymptotics of rt+k and of rt+1 + ... + rt+k. These are relevant for the estimation of Value at Risk and other measures of .financial risk in the context of GARCH-models.
Updated on: 03/19/2019 16:54