Credit Risk Assessment in Commercial Banks Based on Support Vector Machines

2006 ◽  
Author(s):  
Wei Sun ◽  
Chen-guang Yang ◽  
Jian-xun Qi
Keyword(s):  
Risk Assessment ◽  
Support Vector Machines ◽  
Credit Risk ◽  
Commercial Banks ◽  
Support Vector ◽  
Credit Risk Assessment ◽  
Vector Machines

scholarly journals Improved forecast assessment for the expected credit losses in credit risk monitoring in commercial banks in the context of international and Russian practices

2020 ◽  
Vol 15 (3) ◽  
pp. 445-457
Author(s):  
Elena Vladimirovna Travkina ◽  
Keyword(s):  
Risk Assessment ◽  
Risk Management ◽  
Credit Risk ◽  
Commercial Banks ◽  
Banking Sector ◽  
Assessment Model ◽  
Financial Instruments ◽  
Credit Risk Assessment ◽  
Ifrs 9 ◽  
Credit Losses

Current banking sector’s performance raises the issues connected with the IFRS 9 Financial Instruments driven transformation of the forecast assessment for the expected credit losses during monitoring and credit risk assessment in commercial banks. In this regard, it becomes important to conduct a comprehensive systematization of the existing Russian and international practices for monitoring and evaluating credit risk in commercial banks. The purpose of the study is to develop a comprehensive approach to the use of an effective model for the impairment of expected losses in banking activities. The novelty of the study includes the enhancement of the tools for the forecast assessment of the expected credit losses among the commercial banks’ clients to improve the credit risk management efficiency. The results from the implementation of IFRS 9 Financial Instruments in the banking area show that modern conditions maintain the uncertainty of the long-term impact of the credit risk on the commercial banks’ performance. What is more, a huge amount of additional information gives significant difficulties, which contributes into the sophisticated calculations of the future credit losses of the banks. It has been justified that a forecast assessment model for the expected credit losses of the clients during the monitoring and bank’s credit risk assessment should be based on the collective or individual ground. The efficient application of the expected losses impairment in the banking performance has been described as a fundamental tool to simulate the expected credit losses to provision for impairment. This model has been shown to be determined by the features of the credit activities and bank portfolio, types of its financial tools, sources of the available information, as well as the applied IT systems. The proposed model validation algorithm for the expected impairment losses could reduce the expected credit losses, decrease the volume of the created assessed reserves, as well as improve the overall commercial bank performance efficiency. Theoretically, the study develops the credit losses risk management in the context of the transformations in the global and Russian banking practices. From the perspective of the practical value, the research gives an opportunity to create an efficient forecast assessment model for the expected credit losses of the commercial banks’ clients, this model contributing into the cost effectiveness of the bank’s credit activities. A promising further research is considered to be aimed at developing the tools for the assessment of the commercial banks’ credit activity results in the context of the adopted changes connected with the introduction of IFRS 9 Financial Instruments in the Russian banking sector.

Credit risk assessment for unbalanced datasets based on data mining, artificial neural network and support vector machines

2018 ◽  
Vol 13 (4) ◽  
pp. 932-951 ◽  
Author(s):  
Sihem Khemakhem ◽  
Fatma Ben Said ◽  
Younes Boujelbene
Keyword(s):  
Support Vector Machines ◽  
Credit Risk ◽  
Characteristic Curve ◽  
Confusion Matrix ◽  
Statistical Prediction ◽  
Support Vector ◽  
Sampling Strategies ◽  
Content Type ◽  
Vector Machines ◽  
Artificial Neural

Purpose Credit scoring datasets are generally unbalanced. The number of repaid loans is higher than that of defaulted ones. Therefore, the classification of these data is biased toward the majority class, which practically means that it tends to attribute a mistaken “good borrower” status even to “very risky borrowers”. In addition to the use of statistics and machine learning classifiers, this paper aims to explore the relevance and performance of sampling models combined with statistical prediction and artificial intelligence techniques to predict and quantify the default probability based on real-world credit data. Design/methodology/approach A real database from a Tunisian commercial bank was used and unbalanced data issues were addressed by the random over-sampling (ROS) and synthetic minority over-sampling technique (SMOTE). Performance was evaluated in terms of the confusion matrix and the receiver operating characteristic curve. Findings The results indicated that the combination of intelligent and statistical techniques and re-sampling approaches are promising for the default rate management and provide accurate credit risk estimates. Originality/value This paper empirically investigates the effectiveness of ROS and SMOTE in combination with logistic regression, artificial neural networks and support vector machines. The authors address the role of sampling strategies in the Tunisian credit market and its impact on credit risk. These sampling strategies may help financial institutions to reduce the erroneous classification costs in comparison with the unbalanced original data and may serve as a means for improving the bank’s performance and competitiveness.

Sign in / Sign up

Export Citation Format

Share Document