Search

The failure of banks to correctly analyze the credit worthiness of their customers has devastating consequences. Therefore, the importance of credit scoring in the banking sector has become a major field of research in recent years. There are some methods such as logistic regression, linear regression, discriminant analysis and artificial neural networks for credit scoring. The subject of this research is to evaluate the performance of machine learning and logistic regression models on credit scoring by comparison. In this study, it is aimed to develop a scorecard model in which banks can be exposed to a minimum level of credit risk by comparing the logistic regression and artificial neural network methods which are two of these methods. Although there are studies on the comparison of credit scoring models in the literature, the studies have been conducted through retail portfolios and a sample that covers a maximum of 4 years. Unlike the studies in the literature, this research was conducted through corporate firms and a larger sample than the studies in the literature. The result of the study indicated that artificial neural networks which have higher success than logistic regression on the development sample, saw lower success on the out of sample data. Thus, while artificial neural networks show higher performance, it is concluded that logistic regression provides more consistent results, and it is thought that artificial neural networks can produce more consistent results by optimization of the iteration processes., Bankaların, müşterilerinin kredi değerliliğini doğru bir şekilde analiz etmemeleri yıkıcı sonuçlar doğurmaktadır. Bu nedenle, bankacılık sektöründe kredi skorlamasının önemi son yıllarda büyük bir araştırma alanı haline gelmiştir. Kredi değerliliğinin skorlanması için lojistik regresyon, doğrusal regresyon, diskriminant analizi ve yapay sinir ağları gibi yöntemler mevcuttur. Bu araştırmanın konusu makine öğrenmesi ve lojistik regresyon modellerinin kredi skorlaması modelindeki performanslarınnı kıyaslama yoluyla değerlendirmektir. Bu çalışma ile klasik yöntemlerle yapay sinir ağlarını karşılaştırarak, bankaların kredi riskine en az düzeyde maruz kalabilecekleri bir skorkart modeli geliştirilmesi amaçlanmıştır. Literatürde kredi skorlaması modellerinin kıyaslanmasına ilişkin çalışmalar mevcut olmakla birlikte, çalışmalar perakende portföyler üzerinden ve en fazla 4 yılı kapsayan bir örneklem üzerinden yapılmıştır. Araştırma literatürdeki çalışmalardan farklı olarak kurumsal firmalar üzerinden ve literatürdeki çalışmalara göre daha geniş bir örneklem üzerinden ele alınmıştır. Çalışma sonucunda geliştirme örnekleminde daha yüksek başarı sergileyen yapay sinir ağlarının, örneklem dışı veri seti üzerinde lojistik regresyondan daha düşük bir performans sergilediği görülmüştür. Böylece yapay sinir ağları yüksek performans gösterse de, lojistik regresyonun daha tutarlı sonuçlar verdiği bulgusuna ulaşılmakla birlikte yapay sinir ağlarının iterasyon süreçlerinde optimizasyon yapılması ile daha tutarlı sonuçlar üretebileceği düşünülmektedir.

Supervisor: Dam Hieu Chi, 先端科学技術研究科, 修士（知識科学）

This paper proposes a credit scoring model for the empirical assessment of default risk drivers of shipping bank loans. A unique dataset, consisting of the credit portfolio of a ship-lending bank is used to estimate a logit model with two-way clustered adjusted standard errors, ensuring robust inferences. Industry specific variables, captured through current and expected conditions in the extremely volatile global shipping freight markets, the risk appetite of borrowers–the shipowners – expressed through the chartering policy they follow – and a pricing variable, are shown for the first time to be the important factors explaining default probabilities of bank loans.

O objetivo deste estudo é analisar o cumprimento das dívidas das pequenas e médias empresas em Benguela/Angola e demonstrar que as Técnicas de Credit Scoring são úteis na definição, solvabilidade e cumprimento das dívidas das empresas em Angola, e recorrendo a essas técnicas de gestão de risco de crédito, quanto à tomada de decisão, o risco de incumprimento da dívida das pequenas e médias empresas diminuirá consideravelmente, na ordem de 90% no setor bancário angolano. A atual situação económica e financeira de Angola, devido à queda do preço do petróleo, constitui uma preocupação para o Executivo Angolano, e este conclui que a diversificação da economia é necessária para a estabilidade. Para se diversificar a economia angolana de modo a atingir excelência, é preciso adotar comportamento perseverante, aliado a disciplina, organização, instrução e ao conhecimento. As empresas têm de recorrer a financiamentos bancários para o próprio investimento. É também indispensável assegurar uma boa monitorização e avaliação. O nível de incumprimento bancário verificado, levam à necessidade de implementar e aperfeiçoar modelos de análise de risco de crédito nos bancos. Os resultados dos questionários apresentados aos funcionários bancários apontam que o processo decisório da concessão de créditos é essencialmente intuitivo em Angola, baseando-se na experiência dos analistas de crédito dos bancos. Logo, neste trabalho propõe-se a utilização de modelos de Credit Scoring como técnicas de gestão de risco nas instituições bancárias aplicado às pequenas e médias empresas em Angola. Procura-se ainda debater os impactos que apresentam as características desses modelos, assim como as suas vantagens e desvantagens. The objective of this study is to analyze the compliance with the debts of small and medium enterprises in Benguela / Angola and to demonstrate that the Credit Scoring Techniques are useful in defining, solvency and compliance with the debts of companies in Angola, and using these techniques of management of Credit risk, in terms of decision-making, the risk of defaulting on the debt of small and medium-sized companies will decrease considerably, by around 90% in the Angolan banking sector. The current economic and financial situation in Angola, due to the fall in the price of oil, is a concern for the Angolan Executive and he concludes that the diversification of the economy is necessary for stability. In order to diversify the Angolan economy in order to achieve excellence, it is necessary to adopt persevering behavior, together with discipline, organization, instruction and knowledge. Companies must resort to bank financing for their own investment. It is also essential to ensure good monitoring and evaluation. The verified level of bank default leads to the need to implement and improve models of credit risk analysis in banks. The results of the questionnaires presented to bank officials point out that the decision-making process for lending is essentially intuitive in Angola, based on the experience of banks' credit analysts. Therefore, this paper proposes the use of Credit Scoring models as risk management techniques in banking institutions applied to small and medium enterprises in Angola. It also seeks to discuss the impacts of the characteristics of these models, as well as their advantages and disadvantages.

Submitted by Isabel Gomes (itg@lisboa.ucp.pt) on 2017-11-29T10:58:45Z No. of bitstreams: 1 TFM_Pedro_Silva_Mestrado_Gestão_Católica_Porto.pdf: 3406279 bytes, checksum: 835950f0518dd1d0b3b8be9e541be317 (MD5) Approved for entry into archive by Isabel Gomes (itg@lisboa.ucp.pt) on 2017-11-29T10:59:26Z (GMT) No. of bitstreams: 1 TFM_Pedro_Silva_Mestrado_Gestão_Católica_Porto.pdf: 3406279 bytes, checksum: 835950f0518dd1d0b3b8be9e541be317 (MD5) Made available in DSpace on 2017-11-29T10:59:26Z (GMT). No. of bitstreams: 1 TFM_Pedro_Silva_Mestrado_Gestão_Católica_Porto.pdf: 3406279 bytes, checksum: 835950f0518dd1d0b3b8be9e541be317 (MD5) Previous issue date: 2017-07-17

This paper is devoted to the estimation of the probability of default (PD) as a crucial parameter in risk management, requests for loans, rating estimation, pricing of credit derivatives and many others key fi nancial fi elds. Particularly, in this paper we will estimate the PD of US banks by means of the statistical models, generally known as credit scoring models. First, in theoretical part, we will briefl y introduce the two main categories of credit scoring models, which will be afterwards used in application part – linear discriminant analysis and regression models (logit and probit), including testing the statistical signifi cance of estimated parameters. In the main part of the paper we will work with the sample of almost three hundred US commercial banks which will be separated into two groups (non-default and default) on the basis of historical information. Subsequently, we will stepwise apply the mentioned above scoring models on this sample to derive several models for estimation of PD. Further we will apply these models to the control sample to determine the most appropriate model. Web of Science 22 2 181 163

V magistrski nalogi sem se posvetila obravnavi standardiziranega procesa odobritve potrošniškega kredita, ki ga odlikuje optimalen potek dela, z vključeno avtomatsko odločitvijo. Standardizacija je dejavnost vzpostavljanja usklajenih pravil in določil z namenom, da se doseže optimalna stopnja urejenosti na danem področju. Standardizacija je nujna, ker predstavlja pogoj za avtomatizacijo. Kreditni proces sestoji iz procesa prodaje, procesa obravnave kreditne vloge, procesa spremljave in odplačila. Večina bank ima v svojih zalednih službah dobro organizirano in avtomatizirano tretjo fazo kreditnega procesa, t.j. proces spremljave, v okviru katerega banka izvaja obračun obresti, spremembe obrestnih mer, spremembe anuitet ipd. Druga faza kreditnega procesa, to je obravnava kreditne vloge, ki je bistveno bolj zahtevna in tudi tvegana, ker je povezana s sprejemanjem odločitev, pa se pogosto izvaja ročno in decentralizirano po poslovalnicah. V ročno izvajanih poslovnih procesih so poslovna pravila in navodila v glavah kreditnih referentov, kar povečuje možnost napak in zlorab, prav tako je čas obdelave kreditne vloge predolg, kar zmanjšuje konkurenčnost banke. Za večino bank so krediti največji in najbolj očiten vir kreditnega tveganja. Za zmanjšanje kreditnega tveganja je ključno dobro upravljanje s tveganji. Banke dobro poznajo tehnike upravljanja s kreditnim tveganjem, saj ima bančni sektor dolgo zgodovino izkušenj na tem področju. S kapitalskim sporazumom Basel II, ki daje velik pomen tehnikam za zmanjševanje kreditnih tveganj in s kapitalskimi olajšavami motivira banke k uporabi notranjih bonitetnih sistemov, so znova pridobili na pomenu modeli kreditnega točkovanja. Modeli kreditnega točkovanja napovedujejo verjetnost, da potencialni kreditojemalec bodisi ne bo poravnal svojih obveznosti ali da bo po določenem časovnem obdobju začel zamujati. Kreditna ocena bankam omogoča hitro in avtomatizirano odločanje, kar je glavna prednost modelov kreditnega točkovanja. Z vidika banke je avtomatizacija odločanja zlasti pomembna pri poslih z velikim številom strank in majhno maržo na enoto produkta, kar je značilno za potrošniške kredite. Na drugi strani, z vidika strank, hitre, transparentne in dosledne bančne odločitve pomembno vplivajo na izboljšanje uporabniške izkušnje. V empiričnem delu naloge, smo na osnovi razkritij iz letnih poročil bank za leto 2014, predstavili in analizirati prakso bank v EU na področju upravljanja in merjenja kreditnih tveganj. Iz vsake izmed 28 držav članic EU smo izbrali po 5 največjih bank, glede na višino bilančne vsote in jih analizirali z vidika uporabe modelov kreditnega točkovanja ter uporabe notranjih bonitetnih ocen za določanje potrebnega kapitala. Analiza vzorca bank je pokazala, da 71% bank držav članic, kreditno sposobnost prebivalstva ocenjuje s pomočjo modelov kreditnega točkovanja in da le 20% bank držav članic kredite prebivalstvu odobrava na podlagi standardiziranega in avtomatiziranega procesa odločanja. Glede na velik razkorak med deležem bank, ki uporabljajo modele kreditnega točkovanja in deležem bank, ki potrošniške kredite odobravajo s pomočjo avtomatiziranega odločanja se poraja upravičen pomislek, da banke v letnih poročilih avtomatiziranega odločanja o kreditu niso razkrile. his master thesis I focused on reading a standardized approval process consumer credit, which is distinguished by optimal workflow with integrated automatic decision. Standardization is the activity of establishing harmonized rules and regulations in order to achieve the optimum degree of order in a given field. Standardization is necessary, as a prerequisite for automation. Credit process consists of the sales process, the process of reading the credit application process monitoring and amortization. Most banks have in their back-office well organized and automated the third stage of the loan process, i process monitoring, in which the bank carries out the calculation of interest, changes in interest rates, changes in annuities, etc. The second phase of the loan process, it is considered a credit application which is substantially more challenging and risky because it is linked to the decision making is often done manually and the decentralized offices. To manually implemented business processes are business rules and instructions in the minds of credit officers, which increases the possibility of errors and abuses, as well as the processing time is too long the role of credit, which reduces the competitiveness of the bank. For most banks, the loans biggest and most obvious source of credit risk. Credit risk mitigation is important to have good risk management. Bank of well known techniques of credit risk management, because the banking sector has a long history of experience in this field. The Basel II capital, which attaches great importance to techniques for reducing credit risk and capital incentives to motivate banks to use internal ratings systems are again on a higher level of credit scoring models. Credit scoring models predict the likelihood that a potential borrower either fail to meet its obligations or that after a certain period of time started to arrive late. The credit score enables banks to quickly and automated decision-making, which is the main advantage of credit scoring models. In terms of the Bank's decision-making automation is particularly important in dealings with a large number of clients and a small margin per unit of product, which is typical for consumer credit. On the other hand, from the perspective of customers, prompt, transparent and consistent banking decisions have a significant impact on improving the user experience. In the empirical part of the dissertation, we based on the disclosures of the Bank's annual report for 2014, presented and analyzed the practice of banks in the EU in the field of management and measurement of credit risk. From each of the 28 EU Member States were selected 5 largest banks, depending on the amount of total assets and analyzed in terms of the use of credit scoring models and the use of internal ratings to calculate capital requirements. Analysis of the sample of banks showed that 71% banks of the Member States, the creditworthiness of the population estimated by using credit scoring models and that only 20% of banks of the Member States as retail loans were granted on the basis of a standardized and automated decision-making process. Given the large gap between the proportion of banks using credit scoring models and the share of banks that consumer loans are granted by means of an automated decision-making raises legitimate concerns that banks in the annual reports, automated credit decision did not reveal.

Financial infrastructure is the underlying foundation of a country’s financial system. It is comprised of all institutions, information, technologies, rules, and standards which enable financial intermediation. The quality of a country’s financial infrastructure determines the efficiency of intermediation, the ability of lenders to evaluate risk and of borrowers to obtain credit, insurance, and other financial products at competitive terms. This report covers two dimensions of Jamaica’s financial infrastructure: 1) payments, remittances, and securities settlement systems, and; 2) credit reporting systems. This technical note does not provide a detailed assessment of individual payments system in the form of a Report on Observance of Standards and Codes (ROSC), but uses the framework of international standards for carrying out a detailed analysis of the existing systems in Jamaica, including the Committee on Payment and Settlement Systems (CPSS) and International Organization of Securities Commissions (IOSCO) Principles for Financial Markets Infrastructure (PFMIs), the CPSS General Guidance for National Payment System Development, the CPSS-World Bank General Principles on International Remittance Services (GPs), the World Bank General Principles for credit reporting and related methodologies. The information used in the assessment includes relevant national laws, regulations, rules and procedures governing the systems and other available material.

The Credit Scoring models are quantitative ones usually used by financial institutions in measure and credit risk forecast, owning consolidated application during the credit concession process of these institutions. This work objectivies to evaluate the possibility of Credit Scoring models application in a microcredit institution denominated Fundo Rotativo de Ação da Cidadania – Cred Cidadania. For this purpose, some data relative to a clients' sample Cred Cidadania were collected and used to develop two Credit Scoring model types: one relating to credit approval and another one named behavioural scoring. The statistical technique used in the models construction was logistic regression. The study results demonstrated that Credit Scoring models obtain satisfactory performance when used in the Cred Cidadania microcredit institution credit risk analysis, as supporting instruments to rely this process. The results also indicate that Credit Scoring models application supplies subsidies to the institution, assisting it in the prevention and reduction of its insolvency as in the decrease of its operational costs, two problems that affect its financial sustainability., Los modelos de Credit Scoring son modelos cuantitativos empleados comúnmente por instituciones financieras en la mensuración y previsión del riesgo de crédito, poseyendo uso consolidado en el proceso de concesión de crédito de estas instituciones. El presente trabajo objetivó evaluar la posibilidad de aplicación de modelos Credit Scoring en una institución de micro crédito denominada Fundo Rotativo de Acción de la Ciudadanía – Cred Ciudadanía. Para eso, fueron recogidos datos relativos a una muestra de clientes del Cred Ciudadanía, y estos datos fueron utilizados para desarrollar dos tipos de modelos de Credit Scoring: uno de aprobación de crédito y un otro llamado behavioural scoring (escoraje comportamental). Las técnicas estadísticas empleadas en la construcción de los modelos fueron análisis discriminante y regresión logística. Los resultados del estudio demostraron que los modelos Credit Scoring obtienen desempeño satisfactorio cuando utilizados en el análisis de riesgo de crédito en la institución de micro crédito Cred Ciudadanía, alcanzando un porcentual de clasificación correcta de las clientas de cerca de 80%. Los resultados indican también que lo uso de modelos Credit Scoring suministra subsidios a la institución, auxiliándola en la prevención y reducción del impago y en la disminución de sus costes operacionales, dos problemas que afectan su sostenibilidad financiera., Os modelos de Credit Scoring são modelos quantitativos empregados comumente por instituições financeiras na mensuração e previsão do risco de crédito, possuindo uso consolidado no processo de concessão de crédito destas instituições. O presente trabalho objetivou avaliar a possibilidade de aplicação de modelos Credit Scoring em uma instituição de microcrédito denominada Fundo Rotativo de Ação da Cidadania – Cred Cidadania. Para isso, foram coletados dados relativos a uma amostra de clientes do Cred Cidadania, e estes dados foram utilizados para desenvolver dois tipos de modelos de Credit Scoring: um de aprovação de crédito e um outro chamado behavioural scoring (escoragem comportamental). As técnicas estatísticas empregadas na construção dos modelos foram análise discriminante e regressão logística. Os resultados do estudo demonstraram que os modelos Credit Scoring obtêm desempenho satisfatório quando utilizados na análise de risco de crédito na instituição de microcrédito Cred Cidadania, alcançando um percentual de classificação correta dos clientes de cerca de 80%. Os resultados indicam também que o uso de modelos Credit Scoring fornece subsídios à instituição, auxiliando-a na prevenção e redução da inadimplência e na diminuição dos seus custos operacionais, dois problemas que afetam a sua sustentabilidade financeira.

Mongolia's economy has embarked on a very high, long-term growth trajectory. To realize fully its economic potential, Mongolia needs to build a diversified, efficient and stable financial system, capable of intermediating both on a large scale and in specific market segments. Access to financial services in Mongolia is relatively high when measured by the demographic penetration of branches. The aim of this technical note is to assess the level of access to finance in Mongolia, and especially for Micro, Small and Medium Enterprises (MSMEs), to identify key obstacles to improving access, and to provide recommendations to overcome these obstacles. The technical note is organized as follows. Section one provides a broad overview of the macroeconomic environment and is followed by section two on the status of access to finance in Mongolia. Section three discusses products and market segments. Section four examines the supply of financial services by analyzing the role of key market players. Section five examines the demand for financial services by drawing on enterprise surveys to assess firms perceptions of their access to finance, and analyzes financing conditions for MSMEs. Section six examines obstacles in the regulatory, supervisory framework, and financial infrastructure for access to finance. Section seven describes the main government programs related to access to finance. In conclusion, section eight provides policy recommendations for overcoming obstacles to enhancing access to finance.

Os modelos de Credit Scoring são modelos quantitativos empregados comumente por instituições financeiras na mensuração e previsão do risco de crédito, possuindo uso consolidado no processo de concessão de crédito destas instituições. O presente trabalho objetivou avaliar a possibilidade de aplicação de modelos Credit Scoring em uma instituição de microcrédito denominada Fundo Rotativo de Ação da Cidadania – Cred Cidadania. Para isso, foram coletados dados relativos a uma amostra de clientes do Cred Cidadania, e estes dados foram utilizados para desenvolver dois tipos de modelos de Credit Scoring: um de aprovação de crédito e um outro chamado behavioural scoring (escoragem comportamental). As técnicas estatísticas empregadas na construção dos modelos foram análise discriminante e regressão logística. Os resultados do estudo demonstraram que os modelos Credit Scoring obtêm desempenho satisfatório quando utilizados na análise de risco de crédito na instituição de microcrédito Cred Cidadania, alcançando um percentual de classificação correta dos clientes de cerca de 80%. Os resultados indicam também que o uso de modelos Credit Scoring fornece subsídios à instituição, auxiliando-a na prevenção e redução da inadimplência e na diminuição dos seus custos operacionais, dois problemas que afetam a sua sustentabilidade financeira. The Credit Scoring models are quantitative ones usually used by financial institutions in measure and credit risk forecast, owning consolidated application during the credit concession process of these institutions. This work objectivies to evaluate the possibility of Credit Scoring models application in a microcredit institution denominated Fundo Rotativo de Ação da Cidadania – Cred Cidadania. For this purpose, some data relative to a clients' sample Cred Cidadania were collected and used to develop two Credit Scoring model types: one relating to credit approval and another one named behavioural scoring. The statistical technique used in the models construction was logistic regression. The study results demonstrated that Credit Scoring models obtain satisfactory performance when used in the Cred Cidadania microcredit institution credit risk analysis, as supporting instruments to rely this process. The results also indicate that Credit Scoring models application supplies subsidies to the institution, assisting it in the prevention and reduction of its insolvency as in the decrease of its operational costs, two problems that affect its financial sustainability.

Made available in DSpace on 2015-03-05T19:13:44Z (GMT). No. of bitstreams: 0 Previous issue date: 26 Nenhuma As facilidades na concessão de crédito a pessoas físicas têm aumentado no decorrer dos últimos três anos. Variáveis como redução das taxas de juros, aumento de prazos de pagamentos e empréstimos consignados à folha de pagamento possibilitaram à população em geral acesso a aquisição de bens móveis, imóveis entre outros. Neste contexto, a procura por mecanismos mais robustos de análise de risco de crédito, no sentido de evitar ou reduzir os níveis de inadimplência do setor se tornaram necessários. Este estudo objetiva construir um modelo econométrico para estimar a probabilidade do risco de inadimplência em uma Instituição Privada de Ensino Superior. Utilizando a técnica estatística de regressão logística, o modelo de risco de crédito foi construído com base em uma amostra de alunos (pessoas físicas) matriculados na Universidade Católica de Pelotas, situada em Pelotas/RS. As variáveis explicativas do modelo foram obtidas a partir da aplicação de um questionário socioeconômico, que gerou um rol de 59 variáveis d The facilitation in the credit concession to individuals has increased over the last three years. Variables such as the reduction in the interest taxes, increase in maturity, payroll-attached loans, have provided the population in general, with access to consumption property, buildings among others. In this scenario, the search for stronger tools of credit risk analysis, trying to avoid or at least reduce the default rates in the field, has become necessary. This study aims at elaborating an econometric model to predict the probability of default risk in a Private University. By using the statistical technique of logistical regression, the credit risk model has been built based on a sample of students (individuals) enrolled at “Universidade Católica de Pelotas”, located in Pelotas/RS. The explaining variables of the model have been obtained from a socio-economical questionnaire, which has generated 59 variables from which, only 3 were really relevant: existence of previously negotiated debts, possession of a

Credit scoring is regarded as a core competence of commercial banks during the last few decades. A number of credit scoring models have been developed to evaluate credit risk of new loan applicants and existing loan clients. The main purpose of the present paper is to evaluate credit risk in Egyptian banks using credit scoring models. Three statistical techniques are used: discriminant analysis, probit analysis and logistic regression. The credit scoring task is performed on one bank’s personal loans data-set. The results so far revealed that all proposed models gave a better average correct classification rate than the one currently used. Also both type I and type II errors had been calculated in order to evaluate the misclassification costs. Рейтинг кредитоспроможності вважається основною сферою діяльності комерційних банків протягом останніх двадцяти років. З метою оцінки кредитного ризику нових осіб, що звертаються за позикою, та вже існуючих клієнтів, було розроблено велику кількість моделей оцінки кредитоспроможності. Головна мета даної статті - оцінити кредитний ризик єгипетських банків за допомогою використання моделей рейтингу кредитоспроможності. Використано три статистичні методи: дискримінаційний аналіз, пробіт-аналіз та метод логістичної регресії. Проблема оцінки кредитоспроможності вирішується за допомогою використання бази даних споживацьких кредитів одного банку. В результаті проведення дослідження поки що виявлено, що всі запропоновані моделі дали кращий рівень класифікації, ніж методика, яка на сьогодні використовується. З метою оцінки витрат на неправильну класифікацію обчислено два типи похибок (тип І та тип II).