Search results

Social media data from financial websites contain information related to enterprise credit risk. Mining valuable new features in social media data helps to improve prediction performance. This paper proposes a credit risk prediction framework that integrates social media information to improve listed enterprise credit risk prediction in the supply chain.

The prediction framework includes four stages. First, social media information is obtained through web crawler technology. Second, text sentiment in social media information is mined through natural language processing. Third, text sentiment features are constructed. Finally, the new features are integrated with traditional features as input for models for credit risk prediction. This paper takes Chinese pharmaceutical enterprises as an example to test the prediction framework and obtain relevant management enlightenment.

The prediction framework can improve enterprise credit risk prediction performance. The prediction performance of text sentiment features in social media data is better than that of most traditional features. The time-weighted text sentiment feature has the best prediction performance in mining social media information.

The prediction framework is helpful for the credit decision-making of credit departments and the policy regulation of regulatory departments and is conducive to the sustainable development of enterprises.

The prediction framework can effectively mine social media information and obtain an excellent prediction effect of listed enterprise credit risk in the supply chain.

While the Chinese securities market is booming, the phenomenon of listed companies falling into financial distress is also emerging, which affects the operation and development of enterprises and also jeopardizes the interests of investors. Therefore, it is important to understand how to accurately and reasonably predict the financial distress of enterprises.

In the present study, ensemble feature selection (EFS) and improved stacking were used for financial distress prediction (FDP). Mutual information, analysis of variance (ANOVA), random forest (RF), genetic algorithms, and recursive feature elimination (RFE) were chosen for EFS to select features. Since there may be missing information when feeding the results of the base learner directly into the meta-learner, the features with high importance were fed into the meta-learner together. A screening layer was added to select the meta-learner with better performance. Finally, Optima hyperparameters were used for parameter tuning by the learners.

An empirical study was conducted with a sample of A-share listed companies in China. The F1-score of the model constructed using the features screened by EFS reached 84.55%, representing an improvement of 4.37% compared to the original features. To verify the effectiveness of improved stacking, benchmark model comparison experiments were conducted. Compared to the original stacking model, the accuracy of the improved stacking model was improved by 0.44%, and the F1-score was improved by 0.51%. In addition, the improved stacking model had the highest area under the curve (AUC) value (0.905) among all the compared models.

Compared to previous models, the proposed FDP model has better performance, thus bridging the research gap of feature selection. The present study provides new ideas for stacking improvement research and a reference for subsequent research in this field.

European Parliament adopted a new directive on gender balance in corporate boards when by 2026, companies must employ 40% of the underrepresented sex into non-executive directors or 33% among all directors. Therefore, this study aims to analyze the impact of gender diversity (GD) on board of directors and the shareholders’ structure and their impact on the likelihood of company bankruptcy during the COVID-19 pandemic.

The data sample consists of 1,351 companies for 2019 and 2020, of which 173 were large, 351 medium-sized companies and 827 small companies. Three bankruptcy indicators were tested for each company size, and extreme gradient boosting (XGBoost) and logistic regression models were developed. These models were then cross-validated by a 10-fold approach.

XGBoost models achieved area under curve (AUC) over 98%, which is 25% higher than AUC achieved by logistic regression. Prediction models with GD features performed slightly better than those without them. Furthermore, this study indicates the existence of critical mass between 30% and 50%, which decreases the probability of bankruptcy for small and medium companies. Furthermore, the representation of women in ownership structures above 50% decreases bankruptcy likelihood.

This is a pioneering study to explore GD topics by application of ensembled machine learning methods. Moreover, the study does analyze not only the GD of boards but also shareholders. A highly innovative approach is GD analysis based on company size performed in one study considering the COVID-19 pandemic perspective.

Quantitative methods known as scoring models have been traditionally developed for credit granting decisions using statistical procedures. The purpose of this paper is to develop a non‐parametric credit scoring model for micro enterprises that are not maintaining balance sheets, and without having a track record of performance and other credit‐worthy parameters.

Multilayer perceptron procedure is used to evaluate credit reliability in three classes of risk, i.e. bad risk credit, foreclosed risk credit and good risk credit.

The development of a neural network model for micro enterprises facilitates bankers and financial institutions in credit granting decisions in an automatic manner in the Indian context.

This study applies comprehensive information on parameters of financial package prepared by Indian financial institutions and banks to micro enterprises to design a credit risk model. This model, instead of categorizing borrowers in terms of their “ability to pay”, attempts a solution to the unsolved problem of credit availability to micro enterprises in an Indian context, having no past performance track record.

The purpose of this paper is to propose a debt rating index system for small industrial enterprises that significantly distinguishes the default state. This debt rating system is constructed using the F-test and correlation analysis method, with the small industrial enterprise loans of a Chinese commercial bank as the data sample. This study establishes the weighting principle for the debt scoring model: “the more significant the default state, the larger is the weight.” The debt rating system for small industrial enterprises is constructed based on the standard “the higher the debt rating, the lower is the loss given default.”

In this study, the authors selected indexes that pass the homogeneity of variance test based on the principle that a greater deviation of the default sample’s mean from the whole sample’s mean leads to greater significance in distinguishing the default samples from the non-default samples. The authors removed correlated indexes based on the results of the correlation analysis and constructed a debt rating index system for small industrial enterprises that included 23 indexes.

Among the 23 indexes, the weights of 12 quantitative indexes add up to 0.547, while the weights of the remaining 11 qualitative indexes add up to 0.453. That is, in the debt rating of the small industry enterprises, the financial indexes are not capable of reflecting all the debt situations, and the qualitative indexes play a more important role in debt rating. The weights of indexes “X17 Outstanding loans to all assets ratio” and “X59 Date of the enterprise establishment” are 0.146 and 0.133, respectively; both these are greater than 0.1, and the indexes are ranked first and second, respectively. The weights of indexes “X6 EBIT-to- current liabilities ratio,” “X13 Ratio of capital to fixed” and “X78 Legal dispute number” are between 0.07 and 0.09, these indexes are ranked third to fifth. The weights of indexes “X3 Quick ratio” and “X50 Per capital year-end savings balance of Urban and rural residents” are both 0.013, and these are the lowest ranked indexes.

The data of index i are divided into two categories: default and non-default. A greater deviation in the mean of the default sample from that of the whole sample leads to greater deviation from the non-default sample’s mean as well; thus, the index can easily distinguish the default and the non-default samples. Following this line of thought, the authors select indexes that pass the F-test for the debt rating system that identifies whether or not the sample is default. This avoids the disadvantages of the existing research in which the standard for selecting the index has nothing to do with the default state; further, this presents a new way of debt rating. When the correlation coefficient of two indexes is greater than 0.8, the index with the smaller F-value is removed because of its weaker prediction capacity. This avoids the mistake of eliminating an index that has strong ability to distinguish default and non-default samples. The greater the deviation of the default sample’s mean from the whole sample’s mean, the greater is the capability of the index to distinguish the default state. According to this rule, the authors assign a larger weight to the index that exhibits the ability to identify the default state. This is different from the existing index system, which does not take into account the ability to identify the default state.

An increasing number of investors have begun using financial data to develop optimal investment portfolios; therefore, the public financial data shared in the capital market plays a critical role in credit ratings. These data enable investors to understand the credit levels of debtors from a bank perspective; this facilitates predicting the debtor default rate to efficiently evaluate investment risks. The paper aims to discuss these issues.

A credit rating model can be developed to reduce the risk of adverse selection and moral hazard caused by information asymmetry in the loan market. In this study, a random forest (RF) was used to evaluate financial variables and construct credit rating prediction models. Data-mining techniques, including an RF, decision tree, neural networks, and support vector machine, were used to search for suitable credit rating forecasting methods. The distance to default from the KMV model was then incorporated into the credit rating model as a research variable to increase predictive power of various data-mining techniques. In addition, four-level and nine-level classification were set to investigate the accuracy rates of various models.

The experimental results indicated that applying the RF in the variable feature selection process and developing a forecasting model was the most effective method of predicting credit ratings; the four-level and nine-level feature-selection settings achieved 95.5 and 87.8 percent accuracy rates, respectively, indicating that RF demonstrated outstanding feature selection and forecasting capacity.

The experimental cases were based on financial data from public companies in North America.

Practical implication of this study indicates the most effective financial variables were dividends common/ordinary, cash dividends, volatility assumption, and risk-free rate assumption.

The RF model can be used to perform feature selection and efficiently filter numerous financial variables to obtain crediting rating information instantly.

The purpose of this paper is to develop a credit‐scoring model as an aggregate valuation procedure that integrates various financial and non‐financial factors and thereby improves small to medium‐sized enterprises' (SMEs) knowledge about their default risk.

Using panel data from a representative sample of Portuguese SMEs operating in the food or beverage manufacturing sector, this paper develops a logit scoring model to estimate one‐year predictions of default.

The probability of non‐default in the next year is an increasing function of profitability, liquidity, coverage, and activity and a decreasing function of leverage. Smaller firms and those with just one bank relationship have a higher probability of default. The findings suggest that a main bank has incentives to engage in hold up by increasing margins that ex post are too high.

Because SMEs differ from large corporations in their credit risk (e.g., riskier, lower asset correlations), this study has implications for both banks and supervisory actors. Banks should consider qualitative variables when setting internal systems and procedures to manage credit risk. Supervisory institutions should claim mixed credit ratings to determine regulatory capital requirements.

This paper offers a new model, focused specifically on SMEs, and explores the role of financial and non‐financial factors in determining internal credit risks.

The purpose of this paper is to propose a data-driven model to predict credit risks of actors collaborating within a supply chain finance (SCF) network based on the analysis of their network attributes. This can support applying reverse factoring mechanisms in SCFs.

Based on network science, the network measures of the actors collaborating in the investigated SCF are derived through a social network analysis. Then several supervised machine learning algorithms are applied to predict the credit risks of the actors on the basis of their network level and organizational-level characteristics. For this purpose, a data set from an SCF within an automotive industry in Iran is used.

The findings of the research clearly demonstrate that considering the network attributes of the actors within the prediction models can significantly enhance the accuracy and precision of the models.

The main limitation of this research is to investigate the applicability and effectiveness of the proposed model within a single case.

The proposed model can provide a well-established basis for financial intermediaries in SCFs to make more sophisticated decisions within financial facilitation mechanisms.

This study contributes to the existing literature of credit risk evaluation by considering credit risk as a systematic risk that can be influenced by network measures of collaborating actors. To do so, the paper proposes an approach that considers network characteristics of SCFs as critical attributes to predict credit risk.

Data mining for predicting credit risk is a beneficial tool for financial institutions to evaluate the financial health of companies. However, the ubiquity of selecting parameters and the presence of unbalanced data sets is a very typical problem of this technique. This study aims to provide a new method for evaluating credit risk, taking into account not only financial and non-financial variables, but also the class imbalance.

The most significant financial and non-financial variables were determined to build a credit scoring model and identify the creditworthiness of companies. Moreover, the Synthetic Minority Oversampling Technique was used to solve the problem of class imbalance and improve the performance of the classifier. The artificial neural networks and decision trees were designed to predict default risk.

Results showed that profitability ratios, repayment capacity, solvency, duration of a credit report, guarantees, size of the company, loan number, ownership structure and the corporate banking relationship duration turned out to be the key factors in predicting default. Also, both algorithms were found to be highly sensitive to class imbalance. However, with balanced data, the decision trees displayed higher predictive accuracy for the assessment of credit risk than artificial neural networks.

Classification results depend on the appropriateness of data characteristics and the appropriate analysis algorithm for data sets. The selection of financial and non-financial variables, as well as the resolution of class imbalance allows companies to assess their credit risk successfully.

Credit ratings have become one of the primary references for financial institutions to assess credit risk. Conventional credit rating approaches mainly concentrated on two-class classification (i.e. good or bad credit), which lacks adequate precision to perform credit risk evaluations in practice. In addition, most of previous researches directly focussed on employing various data mining techniques, but rare studies discussed the influence of data preprocessing before classifier construction. The paper aims to discuss these issues.

This study considers nine-class classification (i.e. nine credit risk level) to credit rating prediction. For the development of more accurate classifiers, the paper adopts two-stage analysis, which integrates multiple data preprocessing and supervised learning techniques. Specifically, the first stage applies feature selection, data clustering, and data resampling methods to preprocess the data, and then the second stage utilizes several classification techniques and classifier ensembles to construct prediction models.

The results show that Bagging-DT with data resampling method achieves excellent accuracy (82.96 percent), indicating that the proposed two-stage prediction model is better than conventional one-stage models.

Practical implication of this study can lower credit rating expenses and also allow corporations to gain credit rating information instantly.