MSEs Credit Risk Assessment Model Based on Federated Learning and Feature Selection

Federated learning has been used extensively in business inno-vation scenarios in various industries.This research adopts the federated learning approach for the first time to address the issue of bank-enterprise information asymmetry in the credit assessment scenario.First,this research designs a credit risk assessment model based on federated learning and feature selection for micro and small enterprises(MSEs)using multi-dimensional enterprise data and multi-perspective enterprise information.The proposed model includes four main processes:namely encrypted entity alignment,hybrid feature selection,secure multi-party computation,and global model updating.Secondly,a two-step feature selection algorithm based on wrapper and filter is designed to construct the optimal feature set in multi-source heterogeneous data,which can provide excellent accuracy and interpretability.In addition,a local update screening strategy is proposed to select trustworthy model parameters for aggregation each time to ensure the quality of the global model.The results of the study show that the model error rate is reduced by 6.22%and the recall rate is improved by 11.03%compared to the algorithms commonly used in credit risk research,significantly improving the ability to identify defaulters.Finally,the business operations of commercial banks are used to confirm the potential of the proposed model for real-world implementation.

Prioritizing real estate enterprises based on credit risk assessment:an integrated multi‑criteria group decision support framework

Credit risk assessment involves conducting a fair review and evaluation of an assessed subject’s solvency and creditworthiness.In the context of real estate enterprises,credit risk assessment provides a basis for banks and other financial institutions to choose suitable investment objects.Additionally,it encourages real estate enterprises to abide by market norms and provide reliable information for the standardized management of the real estate industry.However,Chinese real estate companies are hesitant to disclose their actual operating data due to privacy concerns,making subjective evalu-ation approaches inevitable,occupying important roles in accomplishing Chinese real estate enterprise credit risk assessment tasks.To improve the normative and reliability of credit risk assessment for Chinese real estate enterprises,this study proposes an integrated multi-criteria group decision-making approach.First,a credit risk assessment index for Chinese real estate enterprises is established.Then,the proposed framework combines proportional hesitant fuzzy linguistic term sets and preference ranking organization method for enrichment evaluation II methods.This approach is suitable for processing large amounts of data with high uncertainty,which is often the case in credit risk assessment tasks of Chinese real estate enterprises involving massive subjec-tive evaluation information.Finally,the proposed model is validated through a case study accompanied by sensitivity and comparative analyses to verify its rationality and feasibility.This study contributes to the research on credit assessment for Chinese real estate enterprises and provides a revised paradigm for real estate enterprise credit risk assessment.

Understanding Credit Risk in Internet Consumer Finance:An Empirical Analysis with a Focus on the Young Generation

In recent years,internet finance has garnered increasing attention from the public.Online lending,emerging within the framework of Internet finance as a pivotal component,has witnessed substantial growth.While online credit,within the realm of Internet finance,presents numerous advantages over traditional lending,it concurrently exposes a plethora of credit risk issues.This study aims to facilitate the effective utilization of online credit tools by the young generation within the context of Internet finance.Additionally,it seeks to ensure the overall stability of the Internet finance environment and mitigate risks for the youth.Given the significance of understanding credit risk management for college students in the age of internet finance,this paper adopts the logistic model to evaluate credit risk in internet consumer finance and provides pertinent recommendations from the perspective of the young generation.

Credit Risk Model Taking Account of Inflation and Its Contribution to Macroeconomic Discussion on Effect of Inflation on Output Growth

We use Extended Merton model(EMM)for estimating the firm’s credit risks in the presence of inflation.We show quantitatively that inflation is an influential factor making either a benign or adverse effect on the firm’s survival,supporting at the microeconomic level New Keynesian findings of the nonlinear inflation effect on output growth.Lower inflation increasing the firm’s expected rate of return can raise its mean year returns and decrease its default probability.Higher inflation,decreasing the expected rate return,makes the opposite effect.The magnitude of the adverse effect depends on the firm strength:for a steady firm,this effect is small,whereas for a weaker firm,it can be fatal.EMM is the only model taking account of inflation.It can be useful for banks or insurance companies estimating credit risks of commercial borrowers over the debt maturity,and for the firm’s management planning long-term business operations.

A credit risk assessment model based on SVM for small and medium enterprises in supply chain finance

Background:Supply chain finance(SCF)is a series of financial solutions provided by financial institutions to suppliers and customers facing demands on their working capital.As a systematic arrangement,SCF utilizes the authenticity of the trade between(SMEs)and their“counterparties”,which are usually the leading enterprises in their supply chains.Because in these arrangements the leading enterprises are the guarantors for the SMEs,the credit levels of such counterparties are becoming important factors of concern to financial institutions’risk management(i.e.,commercial banks offering SCF services).Thus,these institutions need to assess the credit risks of the SMEs from a view of the supply chain,rather than only assessing an SME’s repayment ability.The aim of this paper is to research credit risk assessment models for SCF.Methods:We establish an index system for credit risk assessment,adopting a view of the supply chain that considers the leading enterprise’s credit status and the relationships developed in the supply chain.Furthermore,We conducted two credit risk assessment models based on support vector machine(SVM)technique and BP neural network respectly.Results:(1)The SCF credit risk assessment index system designed in this paper,which contained supply chain leading enterprise’s credit status and cooperative relationships between SMEs and leading enterprises,can help banks to raise their accuracy on predicting a small and medium enterprise whether default or not.Therefore,more SMEs can obtain loans from banks through SCF.(2)The SCF credit risk assessment model based on SVM is of good generalization ability and robustness,which is more effective than BP neural network assessment model.Hence,Banks can raise the accuracy of credit risk assessment on SMEs by applying the SVM model,which can alleviate credit rationing on SMEs.Conclusions:(1)The SCF credit risk assessment index system can solve the problem of banks incorrectly labeling a creditworthy enterprise as a default enterprise,and thereby improve the credit rating status in the process of SME financing.(2)By analyzing and comparing the empirical results,we find that the SVM assessment model,on evaluating the SME credit risk,is more effective than the BP neural network assessment model.This new assessment model based on SVM can raise the accuracy of classification between good credit and bad credit SMEs.(3)Therefore,the SCF credit risk assessment index system and the assessment model based on SVM,is the optimal combination for commercial banks to use to evaluate SMEs’credit risk.

Harnessing Internet finance with innovative cyber credit management

In the last two decades,Internet technologies,such as cloud computing,mobile communications,social media,and big data analytics,have brought tremendous changes to our society and reshaped the business in various industries.Specifically,the mushrooming innovations in the financial area fertilized by information and communication technologies indicates the advent of the Internet finance era.Applying the exploratory research approach,we investigate major innovative Internet-based financial services and classify them into five categories,as of e-commerce,e-payment,e-money market,online loan services,and digital currencies.Then we propose a market structure of Internet finance extended from the traditional financial market.We claim that credit management is the key issue in the marketplace of Internet finance,characterized by big data analytics,in which cyber credit appears as whole-process,multi-dimensional,and holographic.We further suggest that cyber credit be represented in the form of vector to overcome the limits of traditional single-value measure in cyber credit management.Based on this framework,we raise main research issues in Internet finance from the perspectives of theory,technology,and governance.

Social credit:a comprehensive literature review

To avoid credit fraud,social credit within an economic system has become an increasingly important criterion for the evaluation of economic agent activity and guaranteeing the development of a market economy with minimal supervision costs.This paper provides a comprehensive review of the social credit literature from the perspectives of theoretical foundation,scoring methods,and regulatory mechanisms.The study considers the credit of various economic agents within the social credit system such as countries(or governments),corporations,and individuals and their credit variations in online markets(i.e.,network credit).A historical review of the theoretical(or model)development of economic agents is presented together with significant works and future research directions.Some interesting conclusions are summarized from the literature review.(1)Credit theory studies can be categorized into traditional and emerging schools both focusing on the economic explanation of social credit in conjunction with creation and evolution mechanisms.(2)The most popular credit scoring methods include expert systems,econometric models,artificial intelligence(AI)techniques,and their hybrid forms.Evaluation indexes should vary across different target agents.(3)The most pressing task for regulatory mechanisms that supervise social credit to avoid credit fraud is the establishment of shared credit databases with consistent data standards.

A dynamic credit risk assessment model with data mining techniques:evidence from Iranian banks

Giving loans and issuing credit cards are two of the main concerns of banks in that they include the risks of non-payment.According to the Basel 2 guidelines,banks need to develop their own credit risk assessment systems.Some banks have such systems;nevertheless they have lost a large amount of money simply because the models they used failed to accurately predict customers’defaults.Traditionally,banks have used static models with demographic or static factors to model credit risk patterns.However,economic factors are not independent of political fluctuations,and as the political environment changes,the economic environment evolves with it.This has been especially evident in Iran after the 2008-2016 USA sanctions,as many previously reliable customers became unable to repay their debt(i.e.,became bad customers).Nevertheless,a dynamic model that can accommodate fluctuating politicoeconomic factors has never been developed.In this paper,we propose a model that can accommodate factors associated with politico-economic crises.Human judgement is removed from the customer evaluation process.We used a fuzzy inference system to create a rule base using a set of uncertainty predictors.First,we train an adaptive network-based fuzzy inference system(ANFIS)using monthly data from a customer profile dataset.Then,using the newly defined factors and their underlying rules,a second round of assessment begins in a fuzzy inference system.Thus,we present a model that is both more flexible to politico-economic factors and can yield results that are max compatible with real-life situations.Comparison between the prediction made by proposed model and a real non-performing loan indicates little difference between them.Credit risk specialists also approve the results.The major innovation of this research is producing a table of bad customers on a monthly basis and creating a dynamic model based on the table.The latest created model is used for assessing customers henceforth,so the whole process of customer assessment need not be repeated.We assert that this model is a good substitute for the static models currently in use as it can outperform traditional models,especially in the face of economic crisis.

Credit risk evaluation using adaptive Lq penalty SVM with Gauss kernel

In order to improve the performance of support vector machine (SVM) applications in the field of credit risk evaluation, an adaptive Lq SVM model with Gauss kernel (ALqG-SVM) is proposed to evaluate credit risks. The non-adaptive penalty of the object function is extended to (0, 2] to increase classification accuracy. To further improve the generalization performance of the proposed model, the Gauss kernel is introduced, thus the non-linear classification problem can be linearly separated in higher dimensional feature space. Two UCI credit datasets and a real life credit dataset from a US major commercial bank are used to check the efficiency of this model. Compared with other popular methods, satisfactory results are obtained through a novel method in the area of credit risk evaluation. So the new model is an excellent choice.

Research on the Model of Household Credit Risk Evaluation of Rural Microfinance

Since rural microfinance is a credit which grants loans without collateral and guarantees to farmers,it is considerably important to evaluate and control the household credit risk.Through establishing the evaluation index system and then using catastrophe progression theory,three common types of catastrophe system and the normalization formula,we get the comprehensive evaluation.Finally,we take the empirical test and the result shows that this method is simpler and more objective which can be used by the credit cooperatives to decide whether to authorize the loans.

The Credit Risk Assessment Model of Internet Supply Chain Finance: Multi-Criteria Decision-Making Model with the Principle of Variable Weight

The characteristics of the financing model are firstly analyzed when the e-commerce enterprises participate in the supply chain finance. Internet supply chain finance models are divided into three categories with the standard of whether the Electronic commerce enterprises provide funds for small and medium enterprises instead of banks. And then we further study the financing process and the functions of the e-commerce platform with specific examples. Finally, combined with the characteristics of the supply chain finance model, we set up a small and medium enterprises credit evaluation model based on the principle of variable weight with its dynamic data. At the same time, a multi time points and multi indicators decision-making method based on the principle of variable weight is proposed and a specific example is presented. In this paper, the Multi-criteria decision-making model with the principle of variable weight has been used two times. At last, a typical case has been analyzed based on this model with a higher accuracy rate of credit risk assessment.

Credit risk constraint mechanisms in rural financial reform

To research the operating mechanisms of rural financial reform, through setting up a contract model, the constraint roles of reputation and legal intervention on the default risk arising in the operating of the credit union funds are inspected. Analysis indicates that the increase in reputation cost can reduce the probability of union member default behavior and the probability of turning to the law for the credit union funds. Meanwhile, the amount of loans and the interest rates can increase the probability of turning to the law for the credit union funds. Below the marginal values, the penalty mechanisms can reduce the balancing probabilities of member default behavior and turning to the law for the credit union funds, namely, the penalty has some "substitution effect" for turning to the law for the credit union funds.

Measuring the model risk-adjusted performance of machine learning algorithms in credit default prediction

Implementing new machine learning(ML)algorithms for credit default prediction is associated with better predictive performance;however,it also generates new model risks,particularly concerning the supervisory validation process.Recent industry surveys often mention that uncertainty about how supervisors might assess these risks could be a barrier to innovation.In this study,we propose a new framework to quantify model risk-adjustments to compare the performance of several ML methods.To address this challenge,we first harness the internal ratings-based approach to identify up to 13 risk components that we classify into 3 main categories—statistics,technology,and market conduct.Second,to evaluate the importance of each risk category,we collect a series of regulatory documents related to three potential use cases—regulatory capital,credit scoring,or provisioning—and we compute the weight of each category according to the intensity of their mentions,using natural language processing and a risk terminology based on expert knowledge.Finally,we test our framework using popular ML models in credit risk,and a publicly available database,to quantify some proxies of a subset of risk factors that we deem representative.We measure the statistical risk according to the number of hyperparameters and the stability of the predictions.The technological risk is assessed through the transparency of the algorithm and the latency of the ML training method,while the market conduct risk is quantified by the time it takes to run a post hoc technique(SHapley Additive exPlanations)to interpret the output.

Clues from networks:quantifying relational risk for credit risk evaluation of SMEs

Owing to information asymmetry,evaluating the credit risk of small-and mediumsized enterprises(SMEs)is difficult.While previous studies evaluating the credit risk of SMEs have mostly focused on intrinsic risk generated by SMEs,our study considers both intrinsic and relational risks generated by neighbor firms’publicly available risk events.We propose a framework for quantifying relational risk based on publicly available risk events for SMEs’credit risk evaluation.Our proposed framework quantifies relational risk by weighting the impact of publicly available risk events of each firm in an interfirm network—considering the impact of interfirm network type,risk event type,and time dependence of risk events—and combines the relational risk score with financial and demographic features to evaluate SMEs credit risk.Our results reveal that relational risk score significantly improves both discrimination and granting performances of credit risk evaluation of SMEs,providing valuable managerial and practical implications for financial institutions.

Impact of risk management strategies on the credit risk faced by commercial banks of Balochistan

This study aims to identify risk management strategies undertaken by the commercial banks of Balochistan,Pakistan,to mitigate or eliminate credit risk.The findings of the study are significant as commercial banks will understand the effectiveness of various risk management strategies and may apply them for minimizing credit risk.This explanatory study analyses the opinions of the employees of selected commercial banks about which strategies are useful for mitigating credit risk.Quantitative data was collected from 250 employees of commercial banks to perform multiple regression analyses,which were used for the analysis.The results identified four areas of impact on credit risk management(CRM):corporate governance exerts the greatest impact,followed by diversification,which plays a significant role,hedging and,finally,the bank’s Capital Adequacy Ratio.This study highlights these four risk management strategies,which are critical for commercial banks to resolve their credit risk.

The valuation of multi-counterparties CDS with credit rating migration

In this paper,the pricing of a Credit Default Swap(CDS)contract with multiple counterparties is considered.The pricing model takes into account the credit rating migration risk of the reference.It is a new model established under the reduced form framework,where the intensity rates are assumed to have structural styles.We derive from it a non-linear partial differential equation system where both positive and negative correlations of counterparties and the references are considered via a single factor model.Then,an ADI(Alternating Direction Implicit)difference method is used to solve the partial differential equations by iteration.From the numerical results,the comparison of multi-counterparty CDS contract and the standard one are analyzed respectively.Moreover,the impact of default parameters on value of the contracts are discussed.

A high‑dimensionality‑trait‑driven learning paradigm for high dimensional credit classification

To solve the high-dimensionality issue and improve its accuracy in credit risk assessment,a high-dimensionality-trait-driven learning paradigm is proposed for feature extraction and classifier selection.The proposed paradigm consists of three main stages:categorization of high dimensional data,high-dimensionality-trait-driven feature extraction,and high-dimensionality-trait-driven classifier selection.In the first stage,according to the definition of high-dimensionality and the relationship between sample size and feature dimensions,the high-dimensionality traits of credit dataset are further categorized into two types:100<feature dimensions<sample size,and feature dimensions≥sample size.In the second stage,some typical feature extraction methods are tested regarding the two categories of high dimensionality.In the final stage,four types of classifiers are performed to evaluate credit risk considering different high-dimensionality traits.For the purpose of illustration and verification,credit classification experiments are performed on two publicly available credit risk datasets,and the results show that the proposed high-dimensionality-trait-driven learning paradigm for feature extraction and classifier selection is effective in handling high-dimensional credit classification issues and improving credit classification accuracy relative to the benchmark models listed in this study.

Research on Personal Credit Risk Assessment Model Based on Instance-Based Transfer Learning

Personal credit risk assessment is an important part of the development of financial enterprises. Big data credit investigation is an inevitable trend of personal credit risk assessment, but some data are missing and the amount of data is small, so it is difficult to train. At the same time, for different financial platforms, we need to use different models to train according to the characteristics of the current samples, which is time-consuming. In view of these two problems, this paper uses the idea of transfer learning to build a transferable personal credit risk model based on Instance-based Transfer Learning (Instance-based TL). The model balances the weight of the samples in the source domain, and migrates the existing large dataset samples to the target domain of small samples, and finds out the commonness between them. At the same time, we have done a lot of experiments on the selection of base learners, including traditional machine learning algorithms and ensemble learning algorithms, such as decision tree, logistic regression, xgboost and so on. The datasets are from P2P platform and bank, the results show that the AUC value of Instance-based TL is 24% higher than that of the traditional machine learning model, which fully proves that the model in this paper has good application value. The model’s evaluation uses AUC, prediction, recall, F1. These criteria prove that this model has good application value from many aspects. At present, we are trying to apply this model to more fields to improve the robustness and applicability of the model;on the other hand, we are trying to do more in-depth research on domain adaptation to enrich the model.

The Fifth China International Credit and Risk Management Conference

The 5th China International Credit and Risk Man- agement Conference was held on September 24- 26 in Xiamen,Fuiian province.More than 600 enterprises,mainly from the fields of import and export trade,manufacturing and finance,sent their repre-

Research on the Effectiveness of KMV Model in China's Bond Credit Rating Market

In recent years,China's bond market has experienced rapid development,but the pace of credit risk supervision has not kept up.Since 2014,the number of domestic credit bond defaults has increased.In 2016,there were 79 domestic default bonds,with a default amount of up to 40.3 billion Yuan.From the perspective of domestic bond market credit risk supervision and early warning mechanism,rating is not objective,and tracking is not timely also rating methods are backward.Therefore,with the development of big data and other technologies,it is urgent to study credit risk supervision methods suitable for the domestic bond market.On the basis of combing the development of domestic bond market and analyzing the current situation of domestic credit rating,this paper combines the results of theoretical research at home and abroad,the information available in the domestic market,big data mining and automation technology,based on the financial and stock exchange information of listed companies,combined with BS option pricing theory,constructs KMV model.