Abstract

CARO, Norma Patricia; GUARDIOLA, Mariana  and  ORTIZ, Pablo. Classification trees as a tool to predict financial difficulties in Latin American companies through their accounting ratios. Contad. Adm [online]. 2018, vol.63, n.1. ISSN 0186-1042.  https://doi.org/10.22201/fca.24488410e.2018.1148.

The analysis of the accounting information of companies is relevant to detect situations of financial vulnerability. Although this information is usually available to those companies that trade in different stock markets, there is generally no explicit characterization that refers to its financial situation (i.e., in crisis or healthy). In this context, the objective of this work is to analyze the incidence of certain accounting ratios on the financial situation of companies from different Latin American markets: Chile, Peru and Argentina, based on the accounting statements that correspond to the 2001-2011 period. To detect the main accounting ratios that influence companies in crisis, recursive methods were applied: Classification Trees. In this manner, the aim is to have tools that allow predicting the financial status of companies. Profitability proved to be a determining factor in all markets, with a strong incidence in the classification of companies. It was subsequently excluded (to analyze the impact of other ratios) due to the fact that the variables with the biggest impact were different between countries. In the case of Argentina, the size of the company, the working capital, and the asset turnover were the most important predicting variables; in the case of Chile it was the flow of operating funds; and in the case of Peru it was liquidity, followed by asset turnover.

Keywords : classification trees; accounting ratios; financial crisis; Latin American companies; G32; G33; C38.

        · abstract in Spanish     · text in English | Spanish     · English ( pdf ) | Spanish ( pdf )