Análisis de la pérdida esperada por tramos y sus determinantes en una cartera de créditos
Date
2019-05-29
Journal Title
Journal ISSN
Volume Title
Publisher
Pontificia Universidad Católica del Perú
Abstract
La reciente crisis financiera internacional, evidenció la necesidad de fortalecer la
regulación, supervisión y gestión de riesgos del sector financiero. De allí, la necesidad de
crear procedimientos y desarrollar sistemas que permitan gestionar los diferentes riesgos que
enfrenta una entidad financiera. Aguilar, Camargo y Morales (2004) indicaron que “El riesgo
de crédito, es el tipo de riesgo más importante al que debe hacer frente cualquier entidad
financiera” (p.3), por ende el que más pérdida genera a una entidad financiera.
El objetivo del presente estudio es analizar la pérdida esperada (PD) por tramos de
días de mora y sus determinantes mediante matrices de transición para lo cual se consideró
los días de mora de los créditos otorgados a microempresarios al cierre de la cartera de 21
meses (desde mayo 2013 hasta enero 2015). La metodología usada para el análisis de la
pérdida esperada es el test estadístico ANOVA, que permite evaluar la hipótesis de que todas
las medias de las muestras consideradas (en nuestro estudio los tramos de días de mora), son
iguales. Mientras que para analizar los determinantes de la pérdida esperada (Probabilidad de
Incumplimiento (PD), Severidad de la Perdida (LGD) y Exposición al Incumplimiento
(EAD)) se utilizó un Modelo de Regresión Lineal Múltiple, brindándonos información sobre
la relevancia de los determinantes para el cálculo de la pérdida esperada en cada uno de los
tramos estudiados. Finalmente, la conclusión es que el análisis de la pérdida esperada por
tramos de días de mora se puede realizar con un mínimo de 6 tramos, siendo posible pero no
práctico, por la cantidad de información diferenciada, el análisis con un mayor número de
tramos, así también, no se puede afirmar que alguno de los determinantes de la pérdida
esperada no es relevante para su cálculo.
The recent international financial crisis highlighted the need to strengthen the regulation, supervision and risk management of the financial sector. Hence the need for developing procedures and systems to manage the different risks facing a financial institution. Aguilar, Camargo and Morales (2004) indicated that "Credit risk is the most important type of risk that is facing any financial institution" (p3), therefore which generates more loss to a financial institution. The aim of this study is to analyze the expected loss tranches of days past due and their determinants through transition matrices which are considered the days past due loans to microentrepreneurs at the end of the portfolio of 21 months (May 2013 to January 2015). The methodology used for the analysis of the expected loss is the statistical test ANOVA, that allows to evaluate the hypothesis that all the means of the samples considered (in our study tranches of days past due) are equal. While analyzing the determinants of the expected loss (PD, LGD and EAD) a Multiple Linear Regression Model was used, giving us information about the relevance of the determinants for the calculation of the expected loss in each of the sections studied. Finally, the conclusion is that the analysis of the expected loss tranches of days past due can be done with a minimum of 6 sections still possible but not practical, differentiated by the amount of information, analysis with a greater number of tranches, also, we can not affirm that one of the determinants of the expected loss is not relevant to the calculation.
The recent international financial crisis highlighted the need to strengthen the regulation, supervision and risk management of the financial sector. Hence the need for developing procedures and systems to manage the different risks facing a financial institution. Aguilar, Camargo and Morales (2004) indicated that "Credit risk is the most important type of risk that is facing any financial institution" (p3), therefore which generates more loss to a financial institution. The aim of this study is to analyze the expected loss tranches of days past due and their determinants through transition matrices which are considered the days past due loans to microentrepreneurs at the end of the portfolio of 21 months (May 2013 to January 2015). The methodology used for the analysis of the expected loss is the statistical test ANOVA, that allows to evaluate the hypothesis that all the means of the samples considered (in our study tranches of days past due) are equal. While analyzing the determinants of the expected loss (PD, LGD and EAD) a Multiple Linear Regression Model was used, giving us information about the relevance of the determinants for the calculation of the expected loss in each of the sections studied. Finally, the conclusion is that the analysis of the expected loss tranches of days past due can be done with a minimum of 6 sections still possible but not practical, differentiated by the amount of information, analysis with a greater number of tranches, also, we can not affirm that one of the determinants of the expected loss is not relevant to the calculation.
Description
Keywords
Mercadeo--Planificación
Citation
Collections
Endorsement
Review
Supplemented By
Referenced By
Creative Commons license
Except where otherwised noted, this item's license is described as info:eu-repo/semantics/openAccess