Avaliação não-paramétrica de desempenho do setor bancário brasileiro

• Ano de defesa: 2006
• Autor(a) principal: Ghilardi, Wanderlei José
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Santa Maria; BR; Administração; UFSM; Programa de Pós-Graduação em Administração
• Palavras-chave em Português: DEA; Desempenho; Bancos; Performance; Banks; CNPQ::CIENCIAS SOCIAIS APLICADAS::ADMINISTRACAO
• Link de acesso: http://repositorio.ufsm.br/handle/1/4511

In this work economic and financial performance of the fifty biggest banks in Brazil are evaluated by using Data Envelopment Analysis (DEA) method applied to accounting demonstrations and other relevant data. Relative efficiency of each bank comparing to the one with the best performance by consider its return on constant and variable scales is shown. Through a set of variables, a unique index of efficiency was produced in order to determine if a firm is above or below the line of efficiency, which represents the reference of the banking sector. From this study, better information is obtained to be used in decision-making process. The performance of each bank is demonstrated in the study. Critical points and suggestions for improvements are also shown. Comparing to the traditional method, the DEA method is more flexible since performance can be measured by a formula with variables which interested them most, then weighted them according to their importances inside the sector. The DEA models also established an optimization on product/consumption ratio, that is, to increase positive factors (products) and to reduce negative factors (consuming elements). The proposed DEA method is oriented to maximize products using available data provided by the Brazilian Central Bank. In order to obtain results, an efficiency level is determined by extracting proportionality between the capacity of generating receipt (principal product) and the net asset including the numbers of employees (principal consumption).