Análisis de la Administración del Riesgo Crediticio en México para Tarjetas de Crédito

Abstract

ResumenLa necesidad de predecir oportunamente a los clientes malos en créditos revolventes en México ha aumentado, es por ello que se propone una mejora al modelo predictivo de incumplimiento utilizado por la regulación local. Este modelo muestra un mejor análisis de características cualitativas y cuantitativas de créditos consolidados que la metodología utilizada por la CNBV en materia de pérdidas esperadas. Las conclusiones de esta investigación muestran la gran posibilidad de optimizar el modelo vigente, minimizando la creación de provisiones, aumentando la rentabilidad por entidad financiera a nivel nacional, cumpliendo con los supuestos teóricos y requerimientos regulatorios a nivel nacional como internacional en la administración del riesgo crediticio. AbstractThe need to predict timely bad customers of revolving loans in Mexico has increased, so it is proposed an improvement to the predictive model used by local regulation. This proposal shows a better analysis of qualitative and quantitative characteristics of consolidated loans than the methodology used by the CNBV on expected losses. The findings of this research show the great possibility to optimize the current model, minimizing the creation of loan loss provisions, increasing profitability by financial institutions in Mexico, complying the theoretical assumptions and regulatory requirements at national and international level in credit risk management.

Banking innovations and their effect on profitability

“…In Mexico, the research carried out by Trejo-Garcia et al (2014) stands out in which they propose an improvement for the current model used by the CNBV for the selection of credit card users to predict default, minimizing the creation of provisions and increasing the profitability of financial institutions and meeting national and international regulatory requirements. The penetration of this financial instrument in our country compared to other countries is observed in Graphic 2.…”