The parameters of the performance repot: R2,K, UI, SR and so on are based on models that assume a Gaussian distribution of returns. I wonder if such parameters could be calculated by considering another kind of distribution and then compare then or play with them in order to get further statistical analysis.