Measuring systemic risk through statistical combination

Although many different definitions of systemic risks are introduced in the literature, some scholars agree to consider that a measure of systemic risk should take into account the links among the institutions of a network. The first consequence is the proposal of different synthetic indices built on the bases of some indicators, but this leads to use a procedure to summarize these indices in an unidimensional one. In the present work we pay attention to the relations among the financial institutions; in particular, we propose an index combining several peculiar variables in order to rank the financial institutions in the network depending on their risk index. The used variables are those proposed by V-lab. Moreover the combination technique may also be considered to perform nonparametric inference, to treat non gaussian distributions as in the case of indices. So we propose to highlight systemic risk in a network of companies performing a nonparametric test to reveal a sort of heterogeneity behavior; in this case the rankings may also be used to create different behavioral groups.