• Refine Query
  • Source
  • Publication year
  • to
  • Language
  • 1
  • Tagged with
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • 1
  • About
  • The Global ETD Search service is a free service for researchers to find electronic theses and dissertations. This service is provided by the Networked Digital Library of Theses and Dissertations.
    Our metadata is collected from universities around the world. If you manage a university/consortium/country archive and want to be added, details can be found on the NDLTD website.
1

Fatores determinantes na análise de crédito da indústria multinacional agrícola de grande porte

Oliveira, Rodrigo Cavaliere 27 February 2015 (has links)
Made available in DSpace on 2016-03-15T19:32:53Z (GMT). No. of bitstreams: 1 Rodrigo Cavaliere Oliveiraprot.pdf: 2617774 bytes, checksum: a0072ccf4589c966207ba25cecb3c1e5 (MD5) Previous issue date: 2015-02-27 / This master thesis examines the importance of the agricultural sector to the Brazilian economy and the particularities in place in Brazil putting credit risk in perspective, how agricultural multinationals perform their credit analysis, which credit models use and which variables are considered more important for good discrimination between good customers and bad customers, in other words, predict with some degree of accuracy customers who will be performing their payments on time and which ones will default. In this context was tested three statistical methods to confirm the theory for this sector and it was compared the accuracy of the results between them, two parametric techniques, logistic regression and discriminant analysis and a third non-parametric, decision trees - CART. The three methods were suitable, with a good explanatory power, especially decision trees and logistic regression. The qualitative variables showed high explanatory power and important for good credit analysis. Among the quantitative variables, liquidity ratios, debt to equity and average payment period stood out as good discriminatory variables for credit purposes / Este estudo visou analisar a importância do setor agrícola para economia brasileira e as particularidades presentes no Brasil sob a ótica de risco de crédito, como as multinacionais agrícolas efetuam suas análises de crédito, que modelos utilizam e quais variáveis são consideradas mais importantes para uma boa discriminação entre clientes bons e clientes ruins, ou seja, prever com certo grau de acurácia os clientes que serão adimplentes e inadimplentes. Nesse contexto foram testados três modelos estatísticos para confirmar a teoria para esse setor e foram comparados os resultados de acerto entre eles. Duas técnicas paramétricas, regressão logística e análise de discriminante, e uma não paramétrica, árvore de decisão CART. Os três modelos se mostraram adequados, com um bom poder explicativo, com um destaque maior para árvore de decisão e regressão logística. As variáveis qualitativas mostraram alto poder explicativos e importantes para uma boa análise de crédito. Dentre as variáveis quantitativas, índices de liquidez, endividamento e prazo médio de pagamento se destacaram como boas discriminadoras de crédito

Page generated in 0.0816 seconds