COMPRENDER ZETA SCORING

The present paper is an application of the Z-score insolvency model from Edward I. Altman to the credit rating of
national clients in HB Fuller Colombia Ltda. This model was developed in order toassess the financial distress of
current clients and predict the performance of new ones as credit subjects. The model is based on the statistical
technique known as multiple discriminated analyses(mda), and was fed with financial ratios from companies
measured in the period from 2001 to 2005. The historical information used was acquired from the Superintendencia
de Sociedades de Colombia Webpage, and the internal information supplied by HB Fuller Colombia Ltda. The
analysis of data was made through the statistical software Statgraphics Plus 5.1. The statistical analysis of thediscriminated function revealed a good predictive performance of the ratios involved in the bankruptcy model for
HB Fuller Colombia Ltda. The accuracy of classification for the final model in the period2001-2005 was 93.81%,
which means it is valid as an appropriate internal methodology in risk credit valuation.
Key Words: Credit Risk, Z-score Model, Discriminant Analysis, Discriminant Function, EdwardAltman, Financial
Ratios, HB Fuller Colombia Ltda., Lambda Wilks, Canonical Correlation, Backtesting.
Introducción a medición del riesgo financiero ha
cobrado vital importancia durante
las últimasdécadas debido al vertiginoso
crecimiento en el volumen
de clientes, tanto del sector bancario
como del sector real, aunado a los
procesos de globalización de los negocios
que exigen unseguimiento
técnico y riguroso de la exposición al
riesgo asociado a la cartera de los clientes actuales y potenciales
de cualquier entidad.5 Estos controles se convierten cada
vez más en una herramientaindispensable para las organizaciones,
ya que una inadecuada gestión de este riesgo puede
avocarlas a una situación de insolvencia e insostenibilidad
financieras.
Los modelos de calificación por...