1                   In this study we try to explain the inclusion of banks in the WDCI list pro posed by Bloomberg. This list collects a group of more than 100 banking in stitutions which, during the crisis, suffered losses. We explain the probability of being part of the list (to suffer severe or highly severe losses) by their structure and performance. These aspects are represented by 4 variables: ROA, tier1 ratio, number of employees and total assets, referred to the two years preceding the crisis, of a larger sample of more than 400 banks com prehending the banks in and outside the list. By considering the heterogeneity among the banks of the list, an explanation of the probability of highly sever losses is offered by considering the previous variables with the addition of in terbanking assets. By using a probit model we find a confirmation of the new rules, inspired by the Basel 3 Accord and by the Financial Stability Board, requiring a solid patrimonial structure, in particular for the “too big to fail” financial institutions, accompanied by a medium return in order to assure a low probability to suffer losses. Keywords: Bank, Tier1, financial crisis, losses Jel classification: G21, G01 ° University of Urbino, Desp, Urbino, email: alessandra.cepparulo@urbino.it * Centro Europa Ricerche  Cer, Rome, email: a.forte@centroeuroparicerche.it The authors would like to thank Giovanni Pesce for his help in preparing the dataset.