Resumen:
he purpose of this paper is to classify and characterize 64 banks,active as of 2010 in Argentina, by means of robust techniques used oninformation gathered during the period 2001-2010. Based on the strat-egy criteria established in Wang, 2007 and Werbin, 2010, seven variableswere selected. In agreement with bank theory, four ?natural? clusterswere obtained, named ?Personal?, ?Commercial?, ?Typical? and ?Otherbanks?, using robust K-means clustering as implemented in R packageRSKC, Kondo, Salibian-Barrera, and Zamar, 2016. In order to undestandthis grouping, projection pursuit based robust principal component analy-sis, Croux and Ruiz-Gazen, 2005 , was conducted on the whole set showingthat essentially three variables can be attributed the formation of differentclusters. In order to reveal each group inner structure, we used R pack-age mclust to fit a finite Gaussian mixture to the data, selecting the bestmodel through Bayesian Information Criterion from ten possible models.This reveale