Evaluating the repayment probability and value-at-risk of the loan portfolio at Vietcombank - Ho Chi Minh City branch

Abstract

Since 2012, bad debt and credit risk has remained a main barrier for the comprehensive development of commercial banking system. To ensure that Vietnam's banking system operates steadily and sustainably, State Bank of Vietnam has driven growth strategy in the banking sector until 2020 by concentrating in credit risk management. Therefore, the quantification of credit risk under international practices becomes an urgent matter at the commercial banks. The more appropriate credit risk quatification will helps banks not only to select customers and appraise loans efficiently, but also to establish essential credit risk reserves and necessary economic capital to prevent economic disruptions. While the Basel Accord has encouraged commercial banks around the world to develop methods and models to quantify credit risk through Value-at-Risk (VaR) framework, commercial banks in Vietnam mainly measure credit risk based on bad debts and overdue debts, the application of modern methods to quantify credit risk has been only at early trial stages. No Vietnamese bank has been able to officially quantify credit risk for itself. From this reality and as a credit officer at the Bank for Foreign Trade of Vietnam – Ho Chi Minh City Branch (VCB.HCM), the author found it necessary to research, develop and apply the method to quantify credit risk at the VCB.HCM branch. This research will present an overview of credit risk and repayment probability, mainly focusing on analysis of the factors affecting the repayment capacity and constructing models for calculating repayment probability of corporate clients at VCB.HCM. The study will examine the probability of default (PD) of corporate customers at VCB.HCM and apply Monte-carlo simulation in the computer application CDOROMv2.3 of Moody’s to find out the appropriate amount of capital for VCB.HCM to cover for potential credit losses through using VaR. The result of the study has showed that the credit loss that may incur at VCB.HCM at 99.90% level of confidence for its portfolio of corporate customers over the last three years is 13.6%. It is then reasonable to suppose that the model built in this paper is simple to apply, easy to understand, and reasonably affordable. The recommendation is that Vietcombank should invest to adopt this type modelling that can be applied at the branch levels or at the system-wide level of Vietcombank. This can be much more economically efficient and more customizable than the commercially available software that can cost millions of dollars at other commercial banks.