Credit Scoring

The main purpose of this paper is to study the problem created by the lack of information about the credit history of some debtors in the databases used to develop credit scoring models and the use of information about behavior compiled by a credit risk register as a potential solution to the problem. The paper analyzes two problems: (i) the need to provide a credit risk estimation of debtors whose behavior is unknown (because they are deleted from the databases without indication of the reason), and (ii) the assessment of the impact of ignoring these data when evaluating the credit risk of entities’ portfolios. The fundamental strategy will be to allow the use of debtors’ credit history in other entities, recorded in the credit risk register.

Dans une économie d'endettement, le système bancaire est peu développé et se caractérise par une forte volatilité. Le risque d'illiquidité des banques. L'enjeu crucial du système bancaire est la création monétaire tout en tenant compte des mesures nécessaires pour faire face au risque d'illiquidité. Ainsi, prévoir le risque de défaillance des emprunteurs nécessite des méthodes de simplification du processus décisionnel dans les accords de crédit. Pour évaluer le risque lié au portefeuille d'engagement, cet article présente deux méthodes statistiques une analyse discriminatoire et une régression logistique.

Several credit-scoring models have been developed using ensemble classifiers in order to improve the accuracy of assessment. However, among the ensemble models, little consideration has been focused on the hyperparameters tuning of base learners, although these are crucial to constructing ensemble models. This study proposes an improved credit scoring model based on the extreme gradient boosting (XGB) classifier using Bayesian hyper-parameters optimization (XGB-BO). The model comprises two steps. Firstly, data pre-processing is utilized to handle missing values and scale the data. Secondly, Bayesian hyper-parameter optimization is applied to tune the hyper-parameters of the XGB classifier and used to train the model. The model is evaluated on four widely public datasets, i.e., the German, Australia, Lending club, and Polish datasets. Several state-of-the-art classification algorithms are implemented for predictive comparison with the proposed method. The results of the proposed model showed promising results, with an improvement in accuracy of 4.10%, 3.03%, and 2.76% on the German, Lending club, and Australian datasets, respectively. The proposed model outperformed commonly used techniques, e.g., decision tree, support vector machine, neural network, logistic regression, random forest, and bagging, according to the evaluation results. The experimental results confirmed that the XGB-BO model is suitable for assessing the creditworthiness of applicants.

The paper compares the models for small business credit scoring developed by logistic regression, neural networks, and CART decision trees on a Croatian bank dataset. The models obtained by all three methodologies were estimated; then validated on the same hold-out sample, and their performance is compared. There is an evident significant difference among the best neural network model, decision tree model, and logistic regression model. The most successful neural network model was obtained by the probabilistic algorithm. The best model extracted the most important features for small business credit scoring from the observed data

Decision tree modelling, as one of data mining techniques, is used for credit scoring of bank customers.The main problem is the construction of decision trees that could classify customers optimally. This study presents a new hybrid mining approach in the design of an effective and appropriate credit scoring model. It is based on genetic algorithm for credit scoring of bank customers in order to offer credit facilities to each class of customers. Genetic algorithm can help banks in credit scoring of customers by selecting appropriate features and building optimum decision trees. The new proposed hybrid classification model is established based on a combination of clustering, feature selection, decision trees, and genetic algorithm techniques. We used clustering and feature selection techniques to pre-process the input samples to construct the decision trees in the credit scoring model. The proposed hybrid model choices and combines the best decision trees based on the optimality criteria. It constructs the final decision tree for credit scoring of customers. Using one credit dataset, results confirm that the classification accuracy of the
proposed hybrid classification model is more than almost the entire classification models that have been compared in this paper. Furthermore, the number of leaves and the size of the constructed decision tree(i.e. complexity) are less, compared with other decision tree models. In this work, one financial dataset was chosen for experiments, including Bank Mellat credit dataset.

This paper presents an alternative technique for financial distress prediction systems. The method is based on a type of neural network, which is called hybrid associative memory with translation. While many different neural network architectures have successfully been used to predict credit risk and corporate failure, the power of associative memories for financial decision-making has not been explored in any depth as yet. The performance of the hybrid associative memory with translation is compared to four traditional neural networks, a support vector machine and a logistic regression model in terms of their prediction capabilities. The experimental results over nine real-life data sets show that the associative memory here proposed constitutes an appropriate solution for bankruptcy and credit risk prediction, performing significantly better than the rest of models under class imbalance and data overlapping conditions in terms of the true positive rate and the geometric mean of true positive and true negative rates.

Este trabalho tem o objetivo de analisar a eficiência do modelo de credit scoring na ação de cross-selling para proporcionar uma maior rentabilidade alinhada ao risco do novo produto. O estudo resultou em 3 cenários de rentabilidade e desempenho. No Cenário 1 sem uso do escoramento apresentou rentabilidade de R$ 0,5 milhões e inadimplência de 16,1%. Nos demais cenários com uso de escores as rentabilidades ultrapassaram R$ 2,3 milhões e inadimplências abaixo de 9%. Os Cenários 2 e 3 apenas com escore de empresas Bureau. O Cenário 4 inclui o modelo Crédit Scoring desenvolvido neste trabalho, apresentou a melhor discriminação entre clientes bons e maus e a maior taxa de aprovação, sendo 75% contra 64% do melhor Bureau. Para isso, utilizou-se de dados fornecido por uma instituição financeira. Utilizando o SPSS e técnicas estatísticas, a análise de Risco Relativo, construção de dummies e a análise de correlação de Spearman, foi gerado o modelo de Regressão Logística Binária, validado com o teste Kolmogorov-Smirnov, a Curva ROC e outros. O modelo de Credit Scoring desenvolvido apresentou resultados satisfatórios quanto a seu poder de classificação dos clientes. A eficácia da Regressão Logística, como ferramenta de predição de performance de crédito, habilita a aplicação da utilização do modelo Credit Scoring pela instituição financeira provedora dos dados para melhorar a rentabilidade e a inadimplência da carteira de clientes com Cartão de Crédito oriundo da carteira de clientes do empréstimo Consignado.

Palavras-Chave: Credit Scoring. Cross-Selling. Risco de Crédito. Rentabilidade.

Previous research on credit scoring that used statistical and intelligent methods was mostly focused on commercial and consumer lending. The main purpose of this paper is to extract important features for credit scoring in small-business lending on a dataset with specific transitional economic conditions using a relatively small dataset. To do this, we compare the accuracy of the best models extracted by different methodologies, such as logistic regression, neural networks (NNs), and CART decision trees. Four different NN algorithms are tested, including backpropagation, radial basis function network, probabilistic and learning vector quantization, by using the forward nonlinear variable selection strategy. Although the test of differences in proportion and McNemar's test do not show a statistically significant difference in the models tested, the probabilistic NN model produces the highest hit rate and the lowest type I error. According to the measures of association, the best NN model also shows the highest degree of association with the data, and it yields the lowest total relative cost of misclassification for all scenarios examined. The best model extracts a set of important features for small-business credit scoring for the observed sample, emphasizing credit programme characteristics, as well as entrepreneur's personal and business characteristics as the most important ones. Copyright © 2005 John Wiley & Sons, Ltd.

China’s Social Credit System (SCS) has captured the imagination and power of big data technology. Launched at the national level in 2014, the system’s aim is to assess the trustworthiness of Chinese citizens in keeping their promises and complying with legal rules, moral norms, and professional and ethical standards. It is essentially a collaborative project between the authorities and the business sectors to rate both individuals and other entities. In certain contexts, each individual will be given a social credit score or rank. Despite the claimed good will to curb escalating dishonesty across societal sectors in China, the worry is that the totality of individuals’ lives will be captured, that citizens will be monitored and that the Orwellian state will become a reality. Individuals risk being reduced to transparent selves before the state in this uneven battle. They are uncertain about what contributes to their social credit scores, how those scores are combined with the state system, and how their data is interpreted and used. In short, the big data-driven SCS is confronting Chinese citizens with major challenges to their privacy and personal data, amongst other interests critical to them.

This article first maps out the background to the construction of China’s big data social laboratory and the SCS, and then summarises the legislative history and evolving concept of social credit. It stresses that apart from the conspicuous SCS policy document introduced by the Chinese central government, pilot legislation has already been implemented at the local levels to regulate the collection and processing of social credit data. The third part critically reviews such local legislation (which often uses the term ‘public credit information’) with reference to personal data protection principles. It also highlights the restrictions on the data subjects’ rights that are placed by the uncoordinated legal framework of personal data and the extra-legal regime of personal archive. It argues that existing legislation and proposed regulations require substantial revisions to mitigate the impacts of the SCS on data privacy. This article hopes to lay the groundwork for further legal study related to social credit and big data in China.

Credit scoring can be defined as a technique that helps credit providers decide Whether to grant credit to consumers or customers. Its increasing importance can be seen from the growing popularity and application of credit scoring in consumer credit. There are advantages not only to construct effective credit scoring models to help improve the bottom-line of credit providers, but also to combine models to yield a better performing combined model. In this paper we described 1) The use of data mining techniques to construct credit scoring models. 2)The combination of credit scoring models to give a superior final model.

This doc contains the first pages of the book. If you want to read the entire book, you could order it in hardback or kindle from the publisher or amazon (but amazon currently charges more than Routledge). Since the hardback is very expensive, I am also offering paperback copies (currently not available on Routledge or Amazon) for €35 (incl. postage). Please e-mail me at manuel.aalbers@kuleuven.be if you are interested. If so, I can send you my payment details.

Due to the financialization of housing in today’s market, housing risks are increas ingly becoming financial risks. Financialization refers to the increasing dominance of financial actors, markets, practices, measurements and narratives.
It also refers to the resulting structural transformation of economies, firms, states and households. This book asserts the centrality of housing to the contemporary capitalist political economy and places housing at the centre of the financialization debate.
A global wall of money is looking for High-Quality Collateral (HQC) investments, and housing is one of the few asset classes considered HQC. This explains why housing is increasingly becoming financialized, but it does not explain its timing, politics and geography. Presenting a diverse range of case studies from the US, the UK, the Netherlands, Germany, Italy and Spain, the chapters in this book include coverage of the role of the state as the driver of financialization processes, and the part played by local and national histories and institutions. This cutting edge volume will pave the way for future research in the area.
Where housing used to be something “local” or “national”, the wo-way coupling of housing to finance has been one crucial element in the recent crisis. It is time to reconsider the financialization of both homeownership and social housing. This book will be of interest to those who study international economics, housing systems, economic geography and financialization.

In financial risk, credit risk management is one of the most important issues in financial decision-making. Reliable credit scoring models are crucial for financial agencies to evaluate credit applications and have been widely studied in the field of machine learning and statistics. Deep learning is a powerful classification tool which is currently an active research area and successfully solves classification problems in many domains. Deep Learning provides training stability, generalization, and scalability with big data. Deep Learning is quickly becoming the algorithm of choice for the highest predictive accuracy. Feature selection is a process of selecting a subset of relevant features, which can decrease the dimensionality, reduce the running time, and improve the accuracy of classifiers. In this study, we constructed a credit scoring model based on deep learning and feature selection to evaluate the applicant's credit score from the applicant's input features. Two public datasets, Australia and German credit ones, have been used to test our method. The experimental results of the real world data showed that the proposed method results in a higher prediction rate than a baseline method for some certain datasets and also shows comparable and sometimes better performance than the feature selection methods widely used in credit scoring.

This study sought to establish how various credit risk management practices affect performance of commercial banks in Nyeri County in Kenya. Even though commercial banks face several types of risks, credit risk stands out as the most severe. Credit risk is the possibility of loss to the lender on non-performing loans. Financial practice as well as theory provides a scientific process of credit risk management in financial institutions. However, lenders still face loan default and consequently this study sought to find out how those practices affect the performance of commercial banks in Nyeri County, Kenya. A census study was conducted where a population of 86 respondents was targeted comprising of branch managers, credit managers and credit officers. The findings of the study were that all commercial banks had a well written credit policy which is strictly and consistently followed. Only few commercial banks conduct a quantitative credit scoring model. In all banks, initial screening is done by credit officer and approval done at different levels depending on the amount. Majority of the banks check post borrowing activities of the borrower. In conclusion, credit risk management has an effect on loan performance amongst commercial banks. Thus, managers should evaluate more accurately the ability to pay back of a customer since the better the screening the better the performance of commercial banks.

Financialization can be characterized as capital switching from the primary, secondary or tertiary circuit to the quaternary circuit of capital. Housing is a central aspect of financialization. The financialization of mortgage markets demands that not just homes but also homeowners become viewed as financially exploitable. It is exemplified by the securitization of mortgage loans, but also by the use of credit scoring and risk-based pricing. In the past century, mortgage markets were transformed from being a 'facilitating market' for homeowners in need of credit to one increasingly facilitating global investment. Since mortgage markets are both local consumer markets and global investment markets, the dynamics of financialization and globalization directly relate homeowners to global investors thereby increasing the volatility in mortgage markets, as the current crisis shows all too well.

Este documento describe la metodología desarrollada por Vapnik (1995), denominada máquinas de vectores de soporte (SVM, por sus siglas en inglés) y realiza dos aplicaciones al caso de clasificación de agentes para el otorgamiento de créditos a partir de sus características. El primer caso de estudio clasifica individuos de un banco alemán. En el segundo caso se pronostica el incumplimiento del pago de créditos comerciales otorgados a empresas colombianas utilizando las características iniciales del crédito. SVM se compara con dos metodologías utilizadas en el análisis de este tipo de problemas, regresión logística y análisis lineal discriminante. Los resultados arrojan un mejor desempeño en la predicción por parte de SVM respecto a las otras dos metodologías.

Financial institutions use credit scorecards for risk management. A scorecard is a data-driven model for predicting default probabilities. Scorecard assessment concentrates on how well a scorecard discriminates good and bad risk. Whether predicted and observed default probabilities agree (i.e., calibration) is an equally important yet often overlooked dimension of scorecard performance. Surprisingly, no attempt has been made to systematically explore different calibration methods and their implications in credit scoring. The goal of the paper is to integrate previous work on probability calibration, to reintroduce available calibration techniques to the credit scoring community, and to empirically examine the extent to which they improve scorecards. More specifically, using real-world credit scoring data, we first develop scorecards using different classifiers, next apply calibration methods to the classifier predictions, and then measure the degree to which they improve calibration. To evaluate performance, we measure the accuracy of predictions in terms of the Brier Score before and after calibration, and employ repeated measures analysis of variance to test for significant differences between group means. Furthermore, we check calibration using reliability plots and decompose the Brier Score to clarify the origin of performance differences across calibrators. The observed results suggest that post-processing scorecard predictions using a calibrator is beneficial. Calibrators improve scorecard calibration while the discriminatory ability remains unaffected. Generalized additive models are particularly suitable for calibrating classifier predictions.

Quantitative methods to assess the creditworthiness of the loan applicants are vital for the profitability and the transparency of the lending business. With the total loan volumes typical for traditional financial institutions, even the slightest improvement in credit scoring models can translate into substantial additional profit. Yet for the regulatory reasons and due to the potential model risk, banks tend to be reluctant to replace the logistic regression as an industrial standard with the new algorithms. This does not stop researchers from examining such new approaches, though. This thesis discusses the potential of the support vector machines, to become an alternative to logistic regression in credit scoring. Using the real-life credit data set obtained from the P2P lending platform Bondora, the scoring models were built to compare the discrimination power of support vector machines against the traditional approach. The results of the comparison were ambiguous. The linear support vector machines performed worse than logistic regression and their training consumed much more time. On the other hand, support vector machines with non-linear kernel performed better than logistic regression and the difference was statistically significant at 95 % level. Despite this success, several factors prevent SVM from the widespread applications in credit scoring, higher training times and lower robustness of the method being two of the major drawbacks. Considering the alternative algorithms which became available in the last 10 years, support vector machines cannot be recommended as a standalone method for credit risk models.

Data mining techniques have numerous applications in credit scoring of customers in the banking field. One of the most popular data mining techniques is the classification method. Previous researches have demonstrated that using the feature selection (FS) algorithms and ensemble classifiers can improve the banks' performance in credit scoring problems. In this domain, the main issue is the simultaneous and the hybrid utilization of several FS and ensemble learning classification algorithms with respect to their parameters setting, in order to achieve a higher performance in the proposed model. As a result, the present paper has developed a hybrid data mining model of feature selection and ensemble learning classification algorithms on the basis of three stages. The first stage, as expected, deals with the data gathering and pre-processing. In the second stage, four FS algorithms are employed, including principal component analysis (PCA), genetic algorithm (GA), information gain ratio, and relief attribute evaluation function. In here, parameters setting of FS methods is based on the classification accuracy resulted from the implementation of the support vector machine (SVM) classification algorithm. After choosing the appropriate model for each selected feature, they are applied to the base and ensemble classification algorithms. In this stage, the best FS algorithm with its parameters setting is indicated for the modeling stage of the proposed model. In the third stage, the classification algorithms are employed for the dataset prepared from each FS algorithm. The results exhibited that in the second stage, PCA algorithm is the best FS algorithm. In the third stage, the classification results showed that the artificial neural network (ANN) adaptive boosting (AdaBoost) method has higher classification accuracy. Ultimately, the paper verified and proposed the hybrid model as an operative and strong model for performing credit scoring.

– The aim of this paper is to present how credit scoring models can be used in financial institutions, in this case in banks, in order to simplify credit lending. Unlike traditional models of credit analysis, scoring models provides valuation based on numerical score who represent clients' possibility to fulfil their obligation. Using credit scoring models, bank can create a numerical snapshot of consumers risk profile. One of the most important characteristic of scoring models is objectivity where two clients with the same characteristics will have the same credit rating. This paper presents some of credit scoring models and the way that financial institutions use them.

This chapter introduces a phenomenon I call 'digital subprime'. This is a domain that represents a rapidly developing frontier in lenders' quest for predictive power involving a growing group of well-funded technology startups who are entering subprime, payday lending markets in various different countries and are lending at high rates of interest to borrowers who often have either poor credit histories or, in some cases, no credit histories at all. In this new variant of consumer credit lending, we see the exploitation of diverse forms of online data, processed through forms of algorithmic analysis, in the attempt to better predict the repayment behaviour of individuals. This data often appears extremely mundane and to have very little to do with the credit product in hand. Data points can range from a user's IP (Internet Protocol) address, to the particular browser they use, to their screen resolution, to information accessed via social media accounts. In examining this new economic phenomenon, the paper presents results from a project that uses digital methods to try to track the trackers being deployed. The chapter seeks to map the terrain of possibility represented by the diverse forms of data that are rendered accessible to lenders, partly as a basis for future research, and partly to highlight key developments in the present and future ontologies of money.

This study provides comprehensive default probability estimation for automobile loans. An extension of the Cox Proportional Hazards model proposed by Vaida and Xu (2000) by incorporating random-effects was used to handle clustered survival data. The estimation procedure was used in the context of credit scoring incorporating term of the financial obligation as a cluster-specific random component. Balancing was performed due to the highly imbalanced nature of the data. Under-sampling was performed to make the censored class less dominant. The estimated model captured dimensions of automobile loan borrowers such as characteristics and payment behavior. Borrowers tend to exhibit correlation in default incidence within the same loan term, but different propensity to default was observed across loan terms. Models estimated using the balanced data exhibit better predictive performance relative to the model estimated using the imbalanced data. The final model can be a basis for developing collection strategies and loan loss provisioning for effective loan management.