Discriminant analysis as a predictive tool of financial difficulties in Brazilian companies in the share market

Abstract

The objective of this study is to verify the economic-financial situation of some companies. For that, it is required the joint analysis of their financial statements in order to predict future conditions which will generate results and honor their commitments. This paper aims at estimating discriminant functions for groups of profitable, intermediary and lossmaking companies listed on the BM&FBOVESPA within 2009 and 2011. The methodological procedures used characterize this study as descriptive, documental and quantitative. The data used were collected from the Economática® database. The companies belonging to the financial sector and services were not considered in the analysis and those which did not show the necessary data were excluded, resulting in a sample of 255 organizations. Data analysis was performed by using the SPSS®software, having as the meeting variable the three groups of companies and as explanatory variables the economic-financial indicators of liquidity, profitability and capital structure. Data analysis identified the existence of separation between groups, pointing to the Debt Breakdown as the variable that best represents this separation. Two functions were created; the first one, which segregated the profitable enterprises from the intermediary ones, showed high ability to demonstrate the differences between the two groups; the second discriminant function accounted for only the residual power for segregation between the intermediary and loss-making companies.