Credit Default Research Articles

Personal bankruptcy prediction using machine learning techniques

It has become crucial to have an early prediction model that provides accurate assurance for users about the financial situation of consumers. Recent studies focused on predicting corporate bankruptcies and credit defaults, not personal bankruptcies. Due to that, this study fills the literature gap by comparing different machine learning algorithms to predict personal bankruptcy. The main objective of the study is to examine the usefulness of machine learning models such as random forest, XGBoost, LightGBM, AdaBoost, CatBoost, and support vector machines in forecasting personal bankruptcy. The research relies on two samples of households (learning and testing) from the Survey of Consumer Finances, which was conducted in the United States. Among the estimated models, CatBoost and XGBoost showed the highest effectiveness. Among the most important variables used in the models are income, refusal to grant credit, delays in the repayment of liabilities, the revolving debt ratio, and the housing debt ratio.

Exploring the Association Between Corporate Financial Credit Risk Management and Corporate Value

In the rapidly changing business environment, corporate financial credit risk management has become a key factor in the stable operation and continuous value enhancement of enterprises. Corporate value is reflected not only in its tangible assets but also in its intangible assets and management capabilities. Among them, financial credit risk management, as an important part of corporate management, is increasingly being paid attention to for its association with corporate value. Financial credit risk management involves the identification, assessment, monitoring, and control of risks related to financial activities, which may stem from market changes, credit defaults, operational errors, and more. Effective financial credit risk management not only helps enterprises reduce potential losses and protect asset safety but also enhances the market reputation and competitiveness of enterprises, thereby increasing their overall value. With the continuous development of financial markets and the strengthening of globalization trends, the financial credit risks faced by enterprises are becoming increasingly complex and varied. Therefore, building a comprehensive financial credit risk management system and enhancing the risk management capabilities of enterprises have become key to achieving sustainable development and creating long-term value.

Early Warning of Financial Risk Management using AGRNN-COA Approach

Financial risk management was used to protect an entity's financial performance and stability, entails detecting, evaluating, and reducing potential risks, such as credit defaults and market swings. It includes tactics including hedging, diversification, and risk limitation. Financial authorities must immediately responsive for financial risk management system that works with the current economic development. This paper proposes a hybrid approach for early pre-alarm warning of financial risk management system and economic development. The proposed hybrid approach is the combined performance of both the Alternating Graph-Regularized Neural Network (AGRNN) and Crayfish Optimization Algorithm (COA).Commonly it is named as AGRNN-COA technique. The major objective of the proposed approach is to give early pre-alarm warning of financial risk management. AGRNN is designed to give early pre-alarm warning of the financial system of the country. The financial risk management from the AGRNN are optimized by using the COA. By then, the proposed model is implemented in the MATLAB/Simulink working platform and the execution is calculated with the present procedures. The proposed method shows better results in all existing like Financial Risk Management Supervised Algorithm (FRM-SVM), Financial Risk Management Deep Neural Network (FRM-DNN) and Financial Risk Management Heap Based Optimization (FRM-HBO). The accuracy level of proposed FRM-AGRNN-COA approach is 98% that is higher than the other existing methods. The specificity and the F-score of the proposed FRM-AGRNN-COA approach is 99% and 97%. The error rate of the proposed FRM-AGRNN-COA approach is 1.8%, which is very less compared to other existing techniques. From the result, it is conclude that the proposed approach based error is less compared to existing techniques.

Entrepreneurial Attitudes in Firms’ Credit Access

Although enterprises make significant contributions to the GDP of countries, bank credit access is one of the biggest obstacles for them to survive. However, businesses can signal their entrepreneurial behaviors that are included in the Theory of Planned Behavior (TPB) to reduce banks’ concerns about their credit default. However, firms can have various entrepreneurial attitudes depending on their international activities that increase their innovativeness, and financial performance enabling credit access. Thus, the impact of TPB on credit access can differ depending on firms’ internationalization status and this is the main of this study to examine. For this purpose, this paper analyzes 1175 firms from various countries. The researchers apply a purposive sampling method to create a research sample and direct an internet-mediated questionnaire to the selected respondents. Binary Logistic Regression analyses have been performed by the researchers to investigate the impact of TPB on credit access. The results show that the components of TPB, personal attitude, personal behavioral control, and subjective norms have different impacts on the credit access of international and domestic companies. Personal attitude and subjective norms negatively affect bank credit access of international and domestic firms, respectively while they do not determine credit access of domestic and international firms, respectively. Moreover, personal behavioral control positively affects credit access of domestic and international companies. The reason for these results might be related to firm-level and country-level characteristics. Relationship lending and the educational and financial support of academicians and policymakers can provide solutions for credit access problems of firms.

Risk Analysis on the Cassava Value Chain in Central Lampung Regency

The Central Lampung Regency is the center of cassava production in Lampung Province, the Indonesia’s largest cassava-producing province. Cassava has much promise; however, various risks arise throughout the value chain. Risks in the agricultural sector, including cassava farming, are unavoidable and severely impact stakeholders and consumers. This study aimed to identify the risks and the most effective mitigation measures for the cassava value chain. This research was conducted using the House of Risk (HOR) method, divided into two phases: HOR 1 and HOR 2. It involved 286 samples encompassing 263 farmers, six traders, three tapioca industries, eight non-tapioca industries, and six financial institutions. The risks were determined through the Supply Chain Operational Reference (SCOR) model. The results unveiled several priority risks in the cassava value chain, consisting of unpredictable weather changes, scarcity of fertilizers, the absence of farming standards, limited capital, delivery technical problems, price fluctuations, negligence of the workforce, default credit, and lack of customer or farmer knowledge leading to confusion. Furthermore, the risk mitigation actions covered preventing the risks of cassava farming; applying appropriate technology; utilizing production, marketing, and financing system assistance; preventing the risks of cassava trading business; implementing a structured market system; developing access to financing, institutions, and markets; developing marketing infrastructure; developing adaptive farming; monitoring during the credit period; socializing risk management to employees; and providing institutional, financing, and production assistance.

Learning from credit default

Utilizing a unique P2P dataset, this study employs the PSM-DID method to explore the learning effect brought about by default events on investors. The findings reveal that investors who experience their first default event demonstrate an improved ability to select a higher-quality project the next time. Notably, this positive effect is more pronounced when facing substantial defaults, as opposed to cases where overdue principal and interest are eventually settled. Investors' initial confidence in defaulted projects contributes to a greater enhancement of their investment skills. Furthermore, the beneficial impacts of defaulted events diminish as investors’ investment experience accumulates.

FAILURE HANDING STRATEGY FOR MICRO PEOPLE BUSINESS CREDIT (KUR)

The smooth operation of the loans released cannot be guaranteed since under certain circumstances, whether intentionally or inadvertently, the debtor breaks his word, making it difficult for Bank BRI Salobulo Kota Palopo to accept the return of the loans that have been given. This study aims to determine the strategy carried out by BRI Salobulo Unit in dealing with KUR Micro credit defaults and the causes and solutions for the handling strategy. The type of research used is qualitative research through interactive data analysis that produces descriptive data, in the form of a fact with the right interpretation to obtain a systematic, factual and accurate picture or painting of facts, characteristics, and the relationship between the phenomenon under study. The results showed that the default on Micro KUR Loans at the BRI Salobulo Bank, Palopo City occurred because the debtor was late in paying his credit bill when it was due. The strategy for handling default loans is adopted by Bank BRI Salobulo Unit through a risk mitigation strategy (preventive) by analyzing customer eligibility with the 5C principle. The rescue strategy (repressive) is carried out by restructuring credit, which is re-launched with the 3R technique (Rescheduling, Reconditioning, Restructuring). If the customer is not cooperative, then the handling is through a settlement strategy (aggressive) by litigation and by selling under the hands. After mapping the SWOT matrix, several strategies are appropriate, and the strategy focuses on the ST strategy.
 

The Effect of Credit Circulation, Loan to Deposit Ratio (LDR), and Interest Rate on Return On Assets (ROA) Due to Non-Performing Loan (NPL) on Credit Unions in Indonesia

<em>Credit Unions or commonly known as CUs are quite developed in Indonesia, especially on the island of Kalimantan. One of the Credit Cooperative Centers that house Primary CU in Indonesia has service areas in 34 Provinces in Indonesia, namely the Credit Union Credit Cooperative Center Indonesia (PUSKOPCUINA). The CU under the auspices of the PUSKOPCUINA movement has grown very well and has become a reference for CUs in Indonesia in terms of development and empowerment. CU's contribution is still not significant in boosting national economic growth, so more comprehensive research is needed. The development of CU in Indonesia is inseparable from the development of the Return On Assets (ROA) contained in the CU. The Kalimantan credit union movement is under the auspices of PUSKOPCUINA which has different regional and cultural characteristics. This is what researchers see as necessary to research the Effect of Outstanding Credit, Loan to Deposit Ratio (LDR), and Loan Interest Rates on Return On Assets (ROA) Due to Default Credit (NPL) at Credit Unions in Indonesia. The research objective was to test and formulate solutions to problems arising from the influence of Outstanding Credit, Loan to Deposit Ratio and Loan Interest Rates on Return On Assets due to Non-Performing Loans. The method used was quantitative research with Path Analysis in the data processing. The research population was the CU movement in Indonesia with a sample of 40 CUs under PUSKOPCUINA. The data used was outstanding credit data, Loan to Deposit ratio, non-performing loans, loan interest rates, and return on assets of 40 CU at PUSKOPCUINA in 2015-2019 with the model used to analyze panel data called the panel data model analyzed using Path Analysis. The results of the study showed that the Loan to Deposit ratio, non-performing loans, and loan interest rates had a significant effect on the return on assets of CUs in Indonesia.</em>

Small and medium-sized enterprises’ time to default: an analysis using an improved mixture cure model with time-varying covariates

The analysis of the survival time of small and medium-sized enterprises (SMEs) in credit default is as important as that of their credit defaults. This analysis is complicated by the fact that not all SMEs ultimately default on their loans, making a standard time-to-event analysis unsuitable. Based on our recently proposed mixture cure model with time-varying covariates, we identify issues in earlier studies and, to update the incidence component of the model, we propose a more flexible method that employs a support vector machine to enhance the exploration of nonlinear covariate effects if SMEs never default. In addition, we consider the time-varying and fixed covariates for the incidence and latency of an event. Our estimation method is tested and verified via a simulation study, which shows that our proposed model outperforms existing models. In our empirical study, an association network was established using fuzzy cognitive maps for feature selection. The results of our analysis reveal that several factors have significant impacts on the survival time to default. The proposed model also gives a favorable performance in the prediction of probability of default.

Machine learning and credit default: a systematic literature review and taxonomy

Machine learning and credit default: a systematic literature review and taxonomy

Moral hazard index for credit risk to SMEs

This article proposes a methodology to calculate the effect of moral hazard on short term credit (working capital) to small and medium-sized enterprises (SMEs). The methodology incorporates four categories of moral hazard ratio defined in a previous study, which are employed to determine probability of default based on a logit model. To this end, a novel Colombian database is used to calculate a moral hazard index that considers the percentages of the odds ratios of the moral hazard variables for positive coefficients on the probabilities of default. The empirical analysis result in an index measuring the impact of moral hazard on odds ratio mainly based on underinvestment moral hazard category in the sample of analyzed companies for the period ranged from 2007 to 2014. • Moral Hazard Index for Credit Risk is proposed to measure the effect of moral hazard. • The original index is based on the analysis of moral hazard ratios on credit default. • Underinvestment category is the most important moral hazard variable in our sample.

The Internal Model of Default Credit for Rural Banks in Indonesia

The urgency of the internal model of default credit for the rural bank in Indonesia is increasing due to the recent acceleration of the increase in Non-Performing Loans at rural banks. This research will formulate an internal model of default credit that can reduce the level of default risk in rural banks based on the explanatory sequential design of mixed research method. The findings of quantitative analysis integration by Chi-Square Analysis, Discriminant Analysis, and Logistics Regression analysis at BPR BKK Pekalongan Regency in 2021 will be explored further by qualitative research. Using nine variables that affect debtors' failure, this study finds that the interest rate is a variable that consistently affects the status of default loans using the integration of 3 analyses and qualitative analysis. The study results indicate that rural banks need to pay more attention to determining credit interest rates when prospective debtors apply for credit. The determination of interest rates is related to compensation for risks faced by rural banks in connection with asymmetric information about the debtor's ability to pay while considering the interest rate determined by the IDIC.

ANALISIS KOMPARASI TINGKAT RISIKO PEMBIAYAAN JUAL BELI DENGAN BAGI HASIL PADA BANK MUAMALAT KCP. BONE

Despite a fairly rapid development of Islamic Banking in Indonesia, but the composition of cost is still imbalance. This is proved by the nomination of skim murabahah (consumptive sale and purchase) compared with skim mudarabah (products division/productive). The cost of mudarabah is the cost of real sector and considered has high risk. Therefore, syariah bank needs to apply risk management and evaluate the level of maximum risk that might happen on the costs spent. The data were analyzed using credit risk method. Credit risk is a model of risk measurement of costs portfolio. This model was introduced by Credit Suisse Financial Products (CSFP) in 1997.This approach wasmaximum loss measurement from default credit portfolio. Data input credit risk are as follow; (exposure of default, default rates, probability of default, and expected loss) after having results of data analysis and then were analyzed using trend analysis to measure the frequency of risk level happened in the period of 2013-2014 and used analysis of hypothesis test one sample T test to obtain the comparison of risk level of sale and purchase of product division. The results of trend analysis express that the frequency of the occurrence of financial risks is more dominant on the sale and purchase contract, it can be seen in June 2013 until December 2014 NPF murabahah continuously improved, it is different with mudarabah (product division) that still run well. While the results of comparative test analysis One sample T test state that the cost of murabahah contract has higher risk measured from the high costs of financing bogged down experienced in this case the collectability category 5 at the end has impact on the level of achievement of profitability of Muamalat banks.

The Default Risk of Bank Customers Based on Embedded Microprocessor Wireless Communication under the Internet Finance Background

Internet finance refers to a new financial business model in which traditional financial institutions and Internet enterprises use Internet technology and information and communication technology to achieve capital financing, payment, investment, and information intermediary services. Embedded microprocessors have more than 32-bit processors with high performance. And it is small in size, light in weight, and low in cost. This article aims to explore the default risk of bank customers based on embedded microprocessor wireless communication under the background of Internet finance and how to control the credit risk of bank customers and help to clarify the development direction of Internet finance. Based on multimedia technology and orderly logistics model, this article designs a customer default credit risk management system. In order to test the performance of the bank customer credit risk management system, a small customer credit evaluation system was created. The experiment proves that the customer credit evaluation system can well realize risk evaluation and risk early warning and promote the sustainable development of the bank. At the same time, the system can be applied to most scenarios in the financial industry.

Assessment of Support Vector Machine performance for default prediction and credit rating

Predicting the creditworthiness of bank customers is a major concern for banking institutions, as modeling the probability of default is a key focus of the Basel regulations. Practitioners propose different default modeling techniques such as linear discriminant analysis, logistic regression, Bayesian approach, and artificial intelligence techniques. The performance of the default prediction is evaluated by the Receiver Operating Characteristic (ROC) curve using three types of kernels, namely, the polynomial kernel, the linear kernel and the Gaussian kernel. To justify the performance of the model, the study compares the prediction of default by the support vector with the logistic regression using data from a portfolio of particular bank customers. The results of this study showed that the model based on the Support Vector Machine approach with the Radial Basis Function kernel, performs better in prediction, compared to the logistic regression model, with a value of the ROC curve equal to 98%, against 71.7% for the logistic regression model. Also, this paper presents the conception of a support vector machine-based rating tool designed to classify bank customers and determine their probability of default. This probability has been computed empirically and represents the proportion of defaulting customers in each class.

Upaya Perlindungan Hukum terhadap Kreditur atas Debitur Wanprestasi dalam Perjanjian Kredit pada Masa Pandemi Covid-19


 Default can occur either due to intentional factors or factors beyond human capabilities, such as what happened to the Amerta Sedana Wisata cooperative, a debtor does not fulfill his agreement by being in arrears for debt payments of up to 10 months. This study aims to reveal legal protection efforts against creditors for defaulting debtors in credit agreements during the Covid-19 pandemic. The research method carried out in this writing is using empirical legal research with a statutory approach (The Statute Approach). The results obtained in this study are forms of legal protection that can be carried out in resolving defaults committed by debtors, namely by using preventive legal protection and repressive legal protection. Legal protection for creditors has been regulated based on the Civil Code Articles 1131 and 1132. In the credit agreement the Amerta Sedana Wisata Savings and Loans Cooperative more often uses non-litigation channels to resolve existing default or non-performing credit problems. Based on the discussion that has been described, the conclusion in this study is that preventive action is an action or effort taken before the occurrence of a dispute and violation of norms. Preventive legal protection is applied by following regulations. Researchers hope that in borrowing money or credit, they should understand correctly how the risks will arise in the future if there is a dispute that arises due to default.

Dynamic modelling of corporate credit ratings and defaults

In this article, we propose a longitudinal multivariate model for binary and ordinal outcomes to describe the dynamic relationship among firm defaults and credit ratings from various raters. The latent probability of default is modelled as a dynamic process which contains additive firm-specific effects, a latent systematic factor representing the business cycle and idiosyncratic observed and unobserved factors. The joint set-up also facilitates the estimation of a bias for each rater which captures changes in the rating standards of the rating agencies. Bayesian estimation techniques are employed to estimate the parameters of interest. Several models are compared based on their out-of-sample prediction ability and we find that the proposed model outperforms simpler specifications. The joint framework is illustrated on a sample of publicly traded US corporates which are rated by at least one of the credit rating agencies S&P, Moody's and Fitch during the period 1995–2014.

Default on a credit prediction using decision tree and ensemble learning techniques

Due to the high importance of credit risk management and the prediction of the default on a credit loan in banks in recent years in most developed countries. This research experimented with three machine learning algorithms which are decision tree and ensemble techniques based on a decision tree-like random forest which is bootstrap aggregation machine learning model and adaptive boosting is a boosting machine learning model. All models were used to perform a binary classification to classify the clients into trusted clients who would pay back the debts who are creditable and a none trusted clients who won’t pay the debts In the time which are who are not creditable. All the models gave nearly the same results based on the selected performance measures which are precision, recall, and f1 measure. There was no significant change when they performed on a real class imbalanced dataset of the default credit card clients of Taiwan.

The Effects of the Brazilian Economic Recession of 2014-2016 on the Default Credit Card and in Private Label Card in the Food Retail

This study explores the effects of the Brazilian recession from 2014 to 2016 on the default of market credit cards and Private Label cards in food retail in a comparative way, through econometric analysis. The research evaluates the credit card default response in the market and, also, Private Label cards, after simulating shocks in macroeconomic variables such as Gross Domestic Product, wages, and unemployment rate, among others, through the application of an autoregressive vector model (VAR model). The data, on a monthly basis, were collected at the Central Bank of Brazil, the Brazilian Institute of Geography and Statistics and at a company in the private label card market. In addition, precedence tests are applied in order to check if there was any causality “in the sense of Granger” on the default of the market credit card and the Private Label card. Among the results found, it was possible to identify that the default rates of the credit card in the market and of the Private Label have different behaviors, from the simulations of the impulse response functions to the identification of the variables that precede them, being that the default of the Private Label card preceded a higher number of variables when compared to the default of credit cards in the market.

Transparency, auditability, and explainability of machine learning models in credit scoring

A major requirement for credit scoring models is to provide a maximally accurate risk prediction. Additionally, regulators demand these models to be transparent and auditable. Thus, in credit scoring, very simple predictive models such as logistic regression or decision trees are still widely used and the superior predictive power of modern machine learning algorithms cannot be fully leveraged. Significant potential is therefore missed, leading to higher reserves or more credit defaults. This article works out different dimensions that have to be considered for making credit scoring models understandable and presents a framework for making “black box” machine learning models transparent, auditable, and explainable. Following this framework, we present an overview of techniques, demonstrate how they can be applied in credit scoring and how results compare to the interpretability of scorecards. A real world case study shows that a comparable degree of interpretability can be achieved while machine learning techniques keep their ability to improve predictive power.