Credit Risk Assessment Methods Research Articles

A hierarchical attention-based feature selection and fusion method for credit risk assessment

A stable financial environment is requisite for the continuous growth of the E-business market, emphasizing the importance of credit risk assessment. Generally, credit risk assessment involves evaluating the creditworthiness of the economic entity and predicting the probability of defaulting on their financial obligations. Existing works mainly focus on exploring more features from various sources to help default prediction, but tend to overlook the feature cost and feature fusion problem. In this paper, we proposed a novel credit risk assessment model based on the hierarchical attention method, characterizing the ability to manage feature acquisition cost and adaptively integrate multi-view features. Particularly, to fulfill the utility of multi-source features, three segmenting standards are proposed to categorize multi-source features into different views according to their characteristics and statistical attributes. Then, a feature-level attention mechanism is proposed to estimate feature importance and guide feature selection, while a view-level attention mechanism is proposed to aggregate view representations and produce default predictions. The proposed method is capable of enhancing crucial features in a hierarchical way and undertaking feature selection to meet feature acquisition cost, thereby boosting the model performance with feature cost awareness. Experimental results show that the proposed method can approximate the best performance with 5% of selected features, and can also achieve a 1.1% AUC improvement compared with the best baseline method over the full feature set, demonstrating our effectiveness of enhancing default prediction performance while reducing feature costs.

Knowledge graph driven credit risk assessment for micro, small and medium-sized enterprises

Micro, Small, and Medium-sized Enterprises (MSMEs) are essential for the growth and development of the country's economy, as they create jobs, generate income, and foster production and innovation. In recent years, credit risk assessment (CRA) has been an essential process used by financial institutions to evaluate the creditworthiness of MSMEs and determine the likelihood of default. Traditionally, CRA has relied on credit scores and financial statements, but with the advent of machine learning (ML) algorithms, lenders have a new tool at their disposal. By and large, ML algorithms are designed to classify borrowers based on their credit history and transactional data while leveraging the entity relationship involved in credit transactions. This study introduces an innovative knowledge graph-driven credit risk assessment model (RGCN-RF) based on the Relational Graph Convolutional Network (RGCN) and Random Forest (RF) algorithm. RGCN is employed to identify topological structures and relationships, which is currently nascent in traditional credit risk assessment methods. RF categorises MSMEs based on the enterprise embedding vector generated from RGCN. Extensive experimentation is conducted to assess model performance utilising the Indian MSMEs database. The balanced accuracy of 92% obtained using the RGCN-RF model demonstrates a considerable advancement over prior techniques in identifying risk-free enterprises.

Credit Risk Assessment of Heavy-Polluting Enterprises: A Wide-ℓp Penalty and Deep Learning Approach

Effective credit risk assessment of heavy-polluting enterprises can achieve a balance between environmental and economic benefits. It requires the consideration of risk indicators for both the carbon information dimension and the compliance dimension. However, as the feature dimensions of the model continue to increase, so does the irrelevant feature or noise. Therefore, we investigate the use of non-integers for regularization from high-dimensional data under the conditions of a large number of irrelevant features. In this paper, a novel Wide-ℓp Penalty and Deep Learning (WPDL) method for credit risk assessment is proposed, which could provide a sparse solution. The Wide-ℓp Penalty component allows feature selection using a linear model with an ℓp Penalty regularization mechanism, where 0 < p ≤ 2. The deep component is a DNN that can generalize indicator features from the credit risk data. The experimental results show that the minimum prediction error occurs at a non-integer ℓp Penalty. Furthermore, the WPDL outperforms other models such as KNN, DT, RF, SVM, MLP, DNN, Gradient Boosting, and Bagging.

Modern Approaches in Credit Risk Modeling: A Literature Review

Abstract In the financial industry, models are pervasive, and their quantity and complexity continue to increase. Constant advancements are made in econometric and statistical theory, but a fast-developing body of rules and regulations governing their use needs modeling specialists to remain vigilant and adaptable. The tendency of these regulations to be ambiguous necessitates that industry professionals and institutions interpret them independently and jointly. This leads in what is referred to as a “industry standard,” or a set of procedures that are recognized among modeling professionals but not necessarily to those outside of the industry. Non-practitioners in the industry may view the modeling department as a “black box” for these reasons. The accurate evaluation of financial credit risk and the forecasting of bankruptcy are crucial to both the economy and society. In recent years, more and more approaches and algorithms have been advanced for this purpose. At this point, it is of the highest concern to investigate the current credit risk assessment methods. In this paper, we review the traditional statistical models and cutting-edge intelligent methods for forecasting financial distress, with a focus on the greatest advances in the academic literature, as the promising trend in this field. Lastly, the paper will conclude with an overview of the evolution of methodologies and conceptual frameworks in credit risk management research, as well as possible future research directions.

The effectiveness of Different Credit Risk Assessment Methods on Loan Performance

The effectiveness of Different Credit Risk Assessment Methods on Loan Performance

Credit risk evaluation using clustering based fuzzy classification method

Credit scoring is a crucial indicator for banks to determine the financial position and the eligibility of a client for credit. In order to assign statistical odds or probabilities to predict the risk of nonpayment in relation to many other factors, the scoring criterion becomes an important issue. The focus of this study is to propose a clustering based fuzzy classification (CBFC) method for credit risk assessment. We aim to illustrate the beneficial use of machine learning (ML) methods whose prediction power is increased by adopting fuzzy theory to calculate the default risk with a better selection of the features contributing to it. An important feature of the CBFC method is that membership values obtained as a result of the fuzzy k-means clustering algorithm are used for the purpose of better capturing the structure of an existing system.An extensive comparison is performed to show how CBFC performs compared to the traditional ones for the datasets having different characteristics in terms of the variable types. Five different real-life datasets are studied to expose the contribution of fuzzy approach on improving the ML use. Our findings show that the proposed CBFC models can produce the promising classification results in credit risk evaluation which aid the practitioners and decision makers for issuance of credit purposes.

Credit Risk Assessment in P2P Lending Using LightGBM and Particle Swarm Optimization

The credit risk evaluation is a vital task in the P2P Lending platform. An effective credit risk assessment method in a P2P lending platform can significantly influence investors' decisions. The machine learning algorithm that can be used to evaluate credit risk as LightGBM, however, the results in evaluating P2P lending need to be improved. The aim of this research is to improve the accuracy of the LightGBM algorithm by combining the Particle Swarm Optimization (PSO) algorithm. The novelty developed in this research is combining LightGBM with PSO for large data from the Lending Club Dataset which can be accessed on Kaggle.com. The highest accuracy also presented satisfactory results with 98.094% of accuracy, 90.514% of Recall, and 97.754% of NPV respectively. The combination of LightGBM and PSO shows better results.

Assessment of associated credit risk in the supply chain based on trade credit risk contagion.

Assessment of associated credit risk in the supply chain is a challenge in current credit risk management practices. This paper proposes a new approach for assessing associated credit risk in the supply chain based on graph theory and fuzzy preference theory. First, we classified the credit risk of firms in the supply chain into two types, namely firms' "own credit risk" and "credit risk contagion"; second, we designed a system of indicators for assessing the credit risks of firms in the supply chain and used fuzzy preference relations to obtain the fuzzy comparison judgment matrix of credit risk assessment indicators, on which basis we constructed the basic model for assessing the own credit risk of firms in the supply chain; third, we established a derivative model for assessing credit risk contagion. On this basis, we carried out a comprehensive assessment of the credit risk of firms in the supply chain by combining the two assessment results, revealing the contagion effect of associated credit risk in the supply chain based on trade credit risk contagion (TCRC). The case study shows that the credit risk assessment method proposed in this paper enables banks to accurately identify the credit risk status of firms in the supply chain, which helps curb the accumulation and outbreak of systemic financial risks.

Financial credit risk assessment of online supply chain in construction industry with a hybrid model chain

Under the influence of COVID-19, although upstream small- and medium-sized enterprises (SMEs) in construction industry chain suffer from high operating costs and tight cash flow problems, their financing demands are even stronger. As an electronic and platform-based comprehensive service, online supply chain finance can ease the financing problems of SMEs in the construction industry. However, it has become an important issue for financial institutions to effectively assess the credit risk in the process of online supply chain financing. In this paper, an online supply chain risk assessment method, which is based on a hybrid model chain, including eXtreme Gradient Boosting (XGBoost), Synthetic Minority Oversampling TEchnique for Nominal and Continuous (SMOTENC), and Random Forest (RF), is proposed to identify and control the credit risk of financial institutions. Specifically, we establish the financial credit risk assessment system with respect to the characteristics of financing enterprises in the construction industry supply chain, including the status of financing enterprises, the status of core enterprises, the operating status of the supply chain, and the status of assets under financing. On the basis of the system, the best index number of the assessment system and the minority samples are obtained by the XGBoost algorithm and the SMOTENC algorithm, respectively. The classification method based on RF is applied to judge the credit risk of financing enterprises in the supply chain of construction industry. In the simulation stage, we take upstream SMEs in the supply chain of construction industry in China as an example for empirical analysis to validate the effectiveness of our proposed method. The credit risk assessment method proposed in this paper has better performance than the commonly used ones in the academic field with an average improvement on assessment accuracy for 6.39% and an average increase of Area Under Curve for 6.95%. Our study provides meaningful exploration on the fund monitoring system of the financing service platform to improve financing efficiency and risk management level.

Commercial Bank Credit Grading Model Using Genetic Optimization Neural Network and Cluster Analysis.

Commercial banks are facing unprecedented credit risk challenges as the financial market becomes more volatile. Based on this, this study proposes and builds a credit risk assessment model for commercial banks based on GANN from the standpoint of commercial banks. In order to provide commercial banks with an effective and dependable credit risk assessment method, the indicators in this study are classified using cluster analysis, and then various representative indicators are chosen using a factor model, which takes into account the comprehensiveness of the information and reduces the complexity of the subsequent empirical analysis. On this basis, the network structure, learning parameters, and learning algorithm of commercial banks' credit risk assessment models are determined. Furthermore, advancements in data preprocessing and genetic operation have been made. According to simulation results, the highest accuracy rate of this method is 94.17 percent, which is higher than the BPNN algorithm 89.46 percent and the immune algorithm 90.14 percent. The optimization algorithm presented in this study improves the convergence speed and search efficiency of traditional algorithms, and the final experimental results show that the scheme is feasible and effective and can be used for commercial bank credit risk assessment.

Developing Credit Risk Assessment Methods to Make loss Provisions for Potential loans

According to Bank of Russia Regulation No. 590-P dated June 28, 2017, Russian banks assess credit risk and make loss provisions for potential loans. Since 01.01.2018, credit institutions have been required to create loss provisions for expected losses in accordance with international standards (IFRS 9). This novation seems relevant due to the lack of a common risk assessment method and the importance of cost optimization on loan provisions. The aim of the study is to improve the credit risk assessment method for making loss provisions for potential loans. The author used the methods of system and logical analysis and synthesis, techniques of high financial calculations, the balance method, the method of financial ratios. When estimating the probability of borrowers’ default, potential credit losses and loan provisions, the author applied actuarial, market, statistical and econometric methods. Based on a Russian bank’s sample data for 2012-2019, the author developed a regression model that establishes the relationship between financial ratios and the default of corporate borrowers — agricultural producers, and checked the significance of the model’s financial ratios. The author divided the borrowers into rating groups by score. The probability of default is the ratio of the number of defaults to the number of borrowers by group. The average default loss for each group depends on the collection / debt ratio in the bank under review. The score of a borrower brings them into a certain rating group, helps calculate the probability of a default and losses in case of default. The calculated expected losses may be of further use when determining loss provisions for potential loans. The author concludes that this method allows assessing risks and making a decision on lending to borrowers — agricultural producers. The expected credit loss approach will allow for more reasonable provisioning, which corresponds to other authors’ findings. Applying this method in a particular bank requires considering the specifics of the composition and structure of the loan portfolio. It is necessary to analyze the impact of the expected credit loss method on the profitability of banks.

A novel dynamic ensemble selection classifier for an imbalanced data set: An application for credit risk assessment

Credit risk assessment is usually regarded as an imbalanced classification task solved by static ensemble classifiers. However, the dynamic ensemble selection (DES) strategy that can select different ensemble classifiers for each query sample is rarely used. Deficiency of the existing DES algorithm in dealing with imbalanced data is the major challenge. In this paper, a novel combined DES model is developed for imbalanced learning problems. To handle the imbalanced data sets, the synthetic minority over-sampling technique is initially used to balance a training set before generating a candidate classifier pool; then, the weighting mechanism of DES-MI (multi-class imbalance) is used to highlight the importance of minority instances when evaluating classifier competences. To further ensure the comprehensive evaluation and right selection of the ensemble classifier, the meta-learning framework of META-DES is used to account for multiple criteria, and the two-step selection strategy of DES-KNN (k-nearest neighbours) is employed to perform a trade-off between the competence and diversity of the classifiers. Our experiments on 15 imbalanced data sets from the KEEL repository show that the proposed model improves the performance of seven known and popular DES algorithms in terms of the area under the curve. Moreover, the type I error rate of the proposed method is lower than that of XGBoost and LightGBM in a real P2P loan data set indicating the efficiency of the proposed method for credit risk assessment.

Improving Loan Loss Provisioning Framework as a Driver of Economic Growth

Various aspects of credit risk have been studied by many researchers. Scientists and practitioners consider different credit risk assessment methods depending on its application, e.g. to determine capital adequacy, to make loss loan provisions, or to estimate its influence on the interest rate. At the same time, there are almost no studies that consider the relationship between loan loss provisioning framework and loan decisions. The study seeks to 1) understand how the practices and procedures of loan loss provisioning impact total gross loans of Russian banks, and 2) identify constraints for insufficient levels of lending and factors that can foster lending.With the use of an econometric model we estimate a quantitative effect of credit portfolio on the growth of loan loss provisions. We base our model on data derived from financial statements of 400 Russian credit institutions between 2014 and 2019. In addition to our empirical model, we analyze statistical data on the development of the Russian banking system and compare the loan loss provisions in Russian and foreign financial organizations. The estimates are based on Russian official statistics and financial statements of banks within and outside Russia. The study reveals that the existing credit risk assessment method that rests on the regulations provided by the Bank of Russia is responsible for excessive loan loss provisions accumulated by Russian banks. This, in turn, affects the volumes of bank loans.In our research we have arrived at the conclusion that the existing loan loss provisioning is excessive. Current loan loss provisions do not correspond to real lending losses. They negatively affect the financial results of credit institutions, resulting in ungrounded refusals to lend, which in turn limits economic growth. These results support the rationale for reinventing the existing framework of loan loss provisioning.

DEBTOR RATING AS A TOOL OF DEBT MANAGEMENT

IFRS 9 “Financial Instruments” introduced a new model of impairment based on expected credit losses, in which the impairment is based on expected credit losses, and the provision for losses is recognized before the credit loss, i.e. companies recognize losses immediately after initial recognition of the financial asset and revise the amount of the provision for expected credit losses at the reporting date. To create a provision for credit losses, IFRS 9 allows using several practical tools, including the rating debtors’ method. However, IFRS 9 does not express a clear opinion on how the expected credit loss for receivables should be calculated. In this regard, in our opinion, it is possible to apply an individual approach to the choice of credit risk assessment method, determining the debtor’s credit rating and the choice of the default probability, and so on. The paper substantiates that the debtors’ rating by the level of corporate default risk is a method that can reliably assess the probable risks. This method uses credit ratings. The paper proposes using the international credit ratings, which will allow a more objective creditworthiness assessment of both foreign and domestic debtors, taking into account macroeconomic factors used by rating agencies to determine the class of credit risk. The article presents the credit rating of Ukraine and changes in the credit rating of Ukraine for 15 years (2004-2019), shows the model of applying the international default probability rates. Two variants of applying this model are offered. Under the first option, the total amount of receivables from the counterparty / group of debtors is multiplied by the percentage of default probability. The second option involves applying the selected ratio according to the credit rating class at the last stage of calculating the expected credit losses by the simplified method. Due to the fact that there is no single approach to choosing the probability of default and everything relies on expert opinion, we propose using the data of the Annual Global Corporate Default And Rating, which is an analysis of market conditions in the world, the corporate defaults overview, the coefficient of bankruptcy probability of economic entities for each of the risk groups. The paper proposes using the annual rate of corporate defaults, as the expected credit losses must be calculated by companies on a regular basis and revalued at least once a year (on the balance sheet date). It is substantiated that the use of the average rate (Average Rate) to assess the probability of default, it is this rate that takes into account the past experience of companies that are in the corresponding zone of default risk for all the periods under consideration.

Loan evaluation in P2P lending based on Random Forest optimized by genetic algorithm with profit score

Loan evaluation is an effective method for credit risk assessment in peer-to-peer (P2P) lending and significantly affects lender investment decisions as well as his/her profits. Besides traditional methods of loan evaluation, machine learning has gained increased attention and has achieved better performance for P2P lending, especially regarding the Random Forest approach. However, the loan evaluation model based on Random Forest aims to improve the overall accuracy, which cannot guarantee that the lender profit is maximized when the overall accuracy is maximized because the profits of each loan are different. To further improve the loan evaluation effect and lender profits, Random Forest optimized using a genetic algorithm with profit score (RFoGAPS) is proposed. First, considering the actual and potential returns and losses, a new profit score is proposed and taken as the optimization objective. Second, the genetic algorithm is used to optimize the combination of decision trees in Random Forest. Then, the dataset of Lending Club is used to evaluate the proposed method. Experimental results show that the RFoGAPS can obtain higher profits for lenders compared with actual profit and traditional methods. Some suggestions are proposed based on experimental results to facilitate the healthy development of P2P lending.

The Research of Third-party B2B Supply Chain Finance Credit Risk Identification & Evaluation System

With the development of Internet technology, many banks in China launched the “Internet+” supply chain financial services. The optimizing of Supply chain financial structure not only solves the problem that small and medium-sized enterprises financing of convenience and high threshold of loan, but also improves the loan efficiency of the supply chain financial services which achieves a new level and the degree of the participants of the supply chain collaboration. Due to Third-party B2B cooperation mode of supply chain financial services emerging in endlessly, Banks take advantage of the cumulative credit data from Third-party B2B e-commerce platform to issue credit loan and dynamically monitor the credit conditions of small and medium-sized enterprises. Which provide a new solution to long-term information asymmetry problem of supply chain finance service. But the credit risk identification and evaluation change a lot in new service mode. According to analyzing the supply chain finance credit risk changes as result of the evolution from supply chain financial to third-party B2B supply chain finance, this paper contract and analyze the traditional and modern credit risk assessment model. On the basis of above contraction and following the principle of credit risk evaluation index system, this paper build the index system of electronic warehouse financing mode and select the grey evaluation model as the third-party B2B supply chain finance business credit risk assessment model. Multi-level grey credit risk assessment method based on Theil disequilibrium index is used to make comprehensive evaluation on enterprise credit. It can get the evaluation value of the enterprise to judge the enterprise credit rating. This method provide more detailed evaluation method of the third-party B2B supply chain finance business for the banks.

Companies Credit Risk Assessment Methods for Investment Decision Making

As the banks have tightened lending requirements, companies look for alternative sources of external funding. One of such is bonds issue. Unfortunately, corporate bonds issue as a source of funding is rare in Lithuania. This occurs because companies face with a lack of information, investors fear to take on credit risk. Credit risk is defined as a borrower’s failure to meet its obligation. Investors, in order to avoid credit risk, have to assess the state of the companies. The goal of the article is to determine the most informative methods of credit risk assessment. The article summarizes corporate lending sources, analyzes corporate default causes and credit risk assessment methods. The study based on the SWOT analysis shows that investors before making an investment decision should evaluate both the business risk,using qualitative method CAMPARI, and the financial risk, using financial ratio analysis.

A Prediction of the Probability of Default of SMEs on the Credit Guarantee Schemes

SMEs (small and medium enterprises) have had a significant impact in improving economic development and social stability in Taiwan. However, due to weaker management constitutions than large enterprises, SMEs usually encounter more financing when running their business. Through the Small and Medium Enterprise Credit Guarantee Fund of Taiwan (SMEG[11]), the government provides a credit guarantee service to SMEs that have development potential, an ordinary operation status, and good credit history. With government help, SMEs can receive financing from banks without worrying about their lack of collateral. Due to SMEG, the rate of SMEs using financing from banks has gradually increased recently; this demonstrated the importance of the credit guarantee scheme to SMEs. However, the guarantee multiple (guaranteed balance/net worth value) of the SMEG almost reaches to a maximum limit of 20 times, indicating that SMEG is undertaking overly high financial risk. In addition to continuously replenishing the fund amount, the SMEG in Taiwan should also actively improve its risk management capability of underwriting loans to support the guarantee balance with its current guarantee capacity. This study uses companies that have applied for special guarantee loans as samples. This study referenced the Private Firm Model (PFM) developed by Moody's KMV company and established a modified model of default probability using a threshold regression model to redefine default point (DP) to examine the adaptability of the credit risk assessment methods for the Taiwan SMEs. Using the receiver operating characteristic curve (ROC curve) to assess the subject companies, the model proposed by this study is believed to provide better identification when used to predict SMEs’ credit quality. The results of this study will hopefully provide a new method of predicting the credit risk of SMEs that do not share their trading information, and assist the credit guarantee scheme in Taiwan in improving its development and management.

A Hybrid Method for Credit Risk Assessment of Bank Customers

A Hybrid Method for Credit Risk Assessment of Bank Customers

To Construct an Individual Credit Risk Assessment Method

To achieve low costs and better accuracy of individual risk assessments, we constructed a practical method based on multiple classifiers. The method includes many singal classifiers, such as decision trees and the cluster analysis. And we tested it empirically. The result shows that the application of the method can achieve better accuracy than any single classifier of it.