Odhad modelů ratingu evropských firem s využitím vícerozměrných statistických metod

Abstract

The aim of this thesis is estimate of rating models of European companies from the consumer industry and retail by multivariate statistical methods. The first part of thesis contains information about the rating, rating history, rating agencies and the process of making the rating. At the end of this chapter is characterization of financial analysis and selected ratios. Second part is a theoretical description of the methodology of selected multivariate statistical methods that are discriminant analysis and logistic regression. The last part of the thesis is an application part that consists of the estimation of rating models using the two-group discriminant analysis, three-group discriminant analysis and binary logistic regression.