Credit risk model for corporate portfolios

Abstract

The credit risk models that became popular in the international banking industry have limited application in Brazil due to the characteristics of its market. The objective of this research is to propose a set of procedures in order to measure the risk of banks' corporate credit portfolios, considering the actual data available in the Brazilian market. In the proposed approach, the losses of the portfolio's companies are modeled individually and then the results are consolidated to obtain the total loss of the portfolio. Using the Monte Carlo simulation, thousands of scenarios are generated in which the future financial situation of the companies belonging to the portfolio is considered. The scenarios generated give rise to possible loss values regarding the companies individually and the portfolio as a whole. The process is illustrated by applying the model to a hypothetical portfolio based on the data of banks' credit portfolios in Brazil. The model generates the loss distribution of the credit portfolio, from which measurements to quantify the risk of the portfolio can be obtained, like the expected loss and unexpected loss, and the economic capital to be allocated by the financial institution can be calculated. The results indicate that the proposed model is an alternative to measure the risk of credit portfolios.