For the multivariate additive model proposed by Hess et al. [2006] which provides an extension of the additive (or incremental loss ratio) method to correlated lines of business, we determine the Gauss–Markov predictors of the calendar year reserves and their mean squared errors of prediction. We also show that the mean squared errors of prediction can be estimated byusing moment estimators for the parameters which determine the correlation structure. The result is of interest with regard to Solvency II, and it is new even in the univariate case considered by Mack [1991].
Dieser Datensatz wurde nicht während einer Tätigkeit an der Universität Mannheim veröffentlicht, dies ist eine Externe Publikation.
Search Authors in