2.17 in the case ß = 0 u

In order to describe the optimal choices, we have to know the stochas tic dominance rules existing among different portfolios (see, among others, Ortobelli (2001)). Thus, consider two stable distributed random variables Wi = S ( Wi, ßWi, mWi), i = 1, 2 with > 1, which have:

1. the same skewness parameters ßW1 = ßW2 2. the same mean mW1 = mW2 = E(Wi) 3. W1 > W2.