An insight on the use of logistic model regression to predict the liquidity risk in small companies

Abstract

This paper aimed to investigate the restrictive factors of working capital management in micro and small business (MSBs) in the city of Viçosa, MG. To perceive the influence of the variable in the probability of liquidity risk, it was used a model of binary logistic regression analysis in which the subordinated variables was built through the identification of working capital problems. Then, the executed binary logistict regression model has obtained a correct prevision force in 87,7% of the cases, presenting a better development to forecasting businesses classified with no liquidity risk. The classification was efficient to show the existence of conditioning variables of liquidity risk by means of significance tests, allowing the preparation of a probability function.