Esse estudo avalia a adequação dos requerimentos absolutos de capital no Brasil para bancos pequenos e grandes separadamente e investiga os requerimentos de capital mínimo para risco de crédito nas diferentes abordagens de Basiléia, em especial o impacto da adoção dos modelos dos ratings internos (IRB) conforme o Edital BCB n. 37/11. Além disso, propõe e avalia a abordagem padronizada dos ratings centralizados, CRBA, para cálculo do Capital Mínimo Exigido (CME) em bancos pequenos e que é baseada na abordagem padronizada em vigor na Europa, mas voltada para dados disponíveis nas Centrais de Risco. A CRBA pertence à família dos modelos internos e busca contribuir com as recentes discussões sobre a reforma regulatória bancária na Europa e nos Estados Unidos. Para os três objetivos mencionados, as metodologias adotadas foram: 1) o Valuet-at-Risk (VaR) não paramétrico de Crédito (CVaR) de Carey (2002) e o paramétrico Creditrisk+ para estimar o capital econômico do Sistema Bancário; seguido da 2) estimação amostral e avaliação do capital regulatório para bancos pequenos e grandes nas abordagens IRB, Basileia 1, abordagem padrão simplificada (SSA); além da 3) avaliação da abordagem proposta nesse estudo, a CRBA. A performance de todas essas abordagens é avaliada frente a cenários de stress ad hoc e durante a Crise de 2008-2009. Os dados utilizados foram exposições de crédito aleatórias colhidas da Nova Central de Risco do Banco Central do Brasil (SCR). Os principais resultados desse estudo são: 1) sugerir um capital regulatório total (Patrimônio de Referência mais provisão) para bancos grandes de 17,5% baseado no CVaR paramétrico de 99,9% e, para pequenos, de 15,31% baseado no CVaR de 99%; 2) sugerir que, de todas as abordagens de Basileia II, o IRB estimado conforme o Edital BCB n. 37/2011 e para as Probabilidade de Default (PDs) calculadas por matrizes de migração do SCR, é o mais conservador; 3) sugerir que a abordagem proposta seja mais sensível ao risco de crédito do que atual brasileira, especialmente no varejo, além de oferecer um nível proteção maior contra choques aleatórios de crédito. Na Crise de 2008-2009, os bancos pequenos e grandes apresentaram respostas muito distintas a choques diversos ou quando os \"estados da economia\" se deterioravam. Os bancos pequenos não atingem o grau de diversificação necessário para minimizar perdas extremas. Por outro lado, do ponto de vista do risco sistêmico, a falência dessas entidades tem impactos muito menores que a de conglomerados bancários de porte. Finalmente, a abordagem proposta CRBA é apresentada como uma alternativa à abordagem atual no Brasil e à abordagem padronizada (SA) nos demais países, em especial na Europa. No Brasil, a CRBA cumpriria o papel de aumentar a sensibilidade a risco de crédito do CME nos bancos pequenos criando incentivos para uma gestão de risco de crédito mais cautelosa e alinhando o nível de capital dos bancos pequenos ao seu risco efetivo. Nos demais países, a CRBA é uma alternativa à abordagem padronizada, que independe da opinião das Agências de Classificação de Risco (ACRs). A CRBA traz dois benefícios: o primeiro de ampliar o escopo dos modelos internos e eliminar a dependência regulatória na opinião das ACRs, diminuindo a oportunidade de arbitragem regulatória com ratings inflacionados e corrigindo incentivos para que as ACRs sejam apenas provedoras de opiniões isentas; e o segundo, de prover os organismos supervisores com um mecanismo de controle (tracking error) sobre a qualidade de gestão de risco dos bancos pequenos por meio das Centrais de Risco. / This work analyses capital requirements adequacy in Brazil both for small and big banks individually and evaluates the minimum capital requirements for credit risk in the different Basel II approaches, especially, the impacts of IRB adoption as stated on Edital BCB n.37/11. Besides, it proposes and evaluates the Centralized Standard Ratings Based Approach (CRBA) to calculate Minimum Capital Requirements (MCR) in small banks. It is inspired in the Basel II Standard Approach (SA) disseminated in Europe, but based on information from the Credit Registers. The CRBA is an internal model approach in line with recent discussions on regulatory reform in Europe and in the US. The methodology to address these three research goals is: the non-parametric credit Value-at-Risk (VaR) or CVaR of Carey(2002) and the parametric Creditrisk+ to estimate the economic capital for the banking system; to evaluate regulatory capital in small and big banks in the IRB, Basel 1 and the Simplified Standard Approach (SSA) on the sample; and to evaluate the CRBA, proposed in this study. The performance of these approaches is confronted with ad hoc stress scenarios and within the Credit Crisis of 2008-2009. The data is comprised of credit exposures available in the Brazilian Credit Register (SCR). This work main results are: 1) to suggest a total regulatory capital (capital and provision) of 17.5% to big banks based on a parametric CVaR (99.9%) and of 15.31% to small banks based on a CVaR (99%); 2) to suggest, based on all Basel II approaches, that the IRB, as stated on Edital BCB n.31/11 and calibrated with the probabilities of default (PD) estimated with transition matrixes from the SCR, is the most conservative approach; 3) to suggest that the proposed approach is more sensitive to credit risk especially in retail and is more effective against stress chocks. Small and big banks behave differently to adverse shocks. The small banks, for instance, have problems diversifying out extreme losses when the \"states of the economy\" deteriorate. On the other hand, considering systemic risk, the bankruptcies of these institutions are much less of a problem than the ones of a big bank. Finally, the CRBA is presented as an alternative to the current approach (SSA) in Brazil and to the Standard Approach (SA) in other countries, specifically in Europe. In Brazil, the CRBA would increase the risk sensitivity of MCR on smaller banks creating incentives to more careful risk management practices and aligning their capital and risk levels. On the other countries, the CRBA is an alternative to the Standard Approach (SA) that is not dependent on Credit Rating Agencies - CRAs\' opinions and brings two additional benefits. First, it is an internal model based approach eliminating regulatory dependence on CRAs\' opinions, minimizing opportunities to regulatory arbitrage with inflated ratings and allowing CRAs to be more of a trustworthy opinion provider. Second, it provides supervisors a tracking error mechanism to evaluate risk management in small banks using Credit Registers.

有實證研究表明，傳統的信貸風險結構模型顯著低估了違約概率以及信貸收益率差。傳統的結構模型有三個可能的問題：1. 因為正態假設，布朗模型在模擬公司資產價值的過程中未能捕捉到極端事件2. 市場微觀結構噪聲扭曲了股票價格所包含信息3. 在到期日前任何時間，標準BS 期權理論方法不足以描述任何破產的可能性。這些問題在過去的文獻中曾分別提及。而在本文中，在不同的信用風險結構模型的基礎上，我們提出了貝葉斯方法去估算公司價值的跳躍擴散過程和微觀結構噪聲。因為企業的資產淨值不能在市場上觀察，本文建議的貝葉斯方法可對隱藏變量和泊松衝擊作出一定的估算，並就後驗分佈進行財務分析。我們應用馬爾可夫鏈蒙特卡羅方法（MCMC）和吉布斯採樣計算每個參數的後驗分佈。以上的做法，允許我們檢查結構性信用風險模型的偏差主要是來自公司價值的分佈、期權理論方法或市場微觀結構噪聲。我們進行模擬研究以確定模型的表現。最後，我們以新興市場的數據實踐我們的模型。 / There is empirical evidence that structural models of credit risk significantly underestimate both the probability of default and credit yield spreads. There are three potential sources of the problems in traditional structural models. First, the Brownian model driving the firm asset value process may fail to capture extreme events because of the normality assumption. Second, the market micro-structure noise in trading may distort the information contained in equity prices within the estimation process. Third, the standard Black and Scholes option-theoretic approach may be inadequate to describe the consequences of bankruptcy at any time before maturity. These potential problems have been handled separately in the literature. In this paper, we propose a Bayesian approach to simultaneously estimate the jump-diffusion firm value process and micro-structure noise from equity prices based on different structural credit risk models. As the firm asset value is not observable but the equity price is, the proposed Bayesian approach is useful in the estimation with hidden variable and Poisson shocks, and produces posterior distributions for financial analysis. We demonstrate the application using the Markov chain Monte Carlo (MCMC) method to obtain the posterior distributions of parameters and latent variable. The proposed approach enables us to check whether the bias of the structural credit risk model is mainly caused by the firm value distribution, the option-theoretic method or the micro-structure noise of the market. A simulation study is conducted to ascertain the performance of our model. We also apply our model to the emerging market data. / Detailed summary in vernacular field only. / Chan, Sau Lung. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2013. / Includes bibliographical references (leaves 62-65). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstracts also in Chinese. / List of Tables --- p.vii / List of Figures --- p.viii / Chapter 1 --- Introduction --- p.1 / Chapter 2 --- Background and Intuition --- p.5 / Chapter 2.1 --- Merton Model with Trading Noise --- p.7 / Chapter 2.2 --- Black-Cox Model with Default Barrier --- p.10 / Chapter 2.3 --- Double Exponential Jump Diffusion Model (KJD Model) --- p.11 / Chapter 2.4 --- Equity Value via Laplace Transforms --- p.13 / Chapter 2.5 --- KJD Model with Trading Noises --- p.15 / Chapter 3 --- Bayesian Analysis --- p.17 / Chapter 3.1 --- Gibbs Sampling and Metropolis-Hastings Method --- p.17 / Chapter 3.2 --- Merton Model with Trading Noises (M1) --- p.19 / Chapter 3.2.1 --- Prior Distribution for M1 --- p.19 / Chapter 3.2.2 --- Posterior Distribution for M1 --- p.20 / Chapter 3.3 --- Merton Model with Default Barrier (M2) --- p.22 / Chapter 3.3.1 --- Prior Distribution for M2 --- p.23 / Chapter 3.3.2 --- Posterior Distribution for M2 --- p.23 / Chapter 3.4 --- KJD Model with Trading Noises (M3) --- p.25 / Chapter 3.4.1 --- Prior Distribution for M3 --- p.26 / Chapter 3.4.2 --- Posterior Distribution for M3 --- p.27 / Chapter 3.5 --- KJD Model with Default Barrier (M4) --- p.33 / Chapter 3.5.1 --- Prior Distribution for M4 --- p.34 / Chapter 3.5.2 --- Posterior Distribution for M4 --- p.35 / Chapter 4 --- Numerical Examples --- p.42 / Chapter 4.1 --- Simulation Analysis --- p.42 / Chapter 4.2 --- Empirical Study --- p.46 / Chapter 4.2.1 --- BEA and DBS, 2003-2004 --- p.46 / Chapter 4.2.2 --- HSBC, 2008-2009 --- p.49 / Chapter 5 --- Conclusion --- p.60 / Bibliography --- p.62

The credit derivatives market has known an incredible development since its advent in the 1990's. Today there is a plethora of credit derivatives going from the simplest ones, credit default swaps (CDS), to more complex ones such as synthetic single-tranche collateralized debt obligations. Valuing this rich panel of products involves modeling credit risk. For this purpose, two main approaches have been explored and proposed since 1976. The first approach is the Structural approach, first proposed by Merton in 1976, following the work of Black-Scholes for pricing stock options. This approach relies in the capital structure of a firm to model its probability of default. The other approach is called the Reduced-form approach or the hazard rate approach. It is pioneered by Duffie, Lando, Jarrow among others. The main thesis in this approach is that default should be modeled as a jump process. The objective of this work is to value Asset-backed Credit default swaps using the hazard rate approach.

This project focuses on risk analysis of corporate bond portfolios. We separate the total risk of the portfolio into three parts, which are market risk, credit risk and liquidity risk. The market risk component is quantified by value-at-risk (VaR) determined by change in yield to maturity of the bond portfolio. For the credit risk component, we calculate default probabilities and losses in the event of default and then compute credit VaR. Next, we define a factor called basis which is the difference between the Credit Default Swap (CDS) spread and its corresponding corporate bond yield spread (z-spread or OAS). We quantify the liquidity risk by using the basis. In addition, we also introduce a Fama-French multi-factor model to analyze factor significance to the corporate bond portfolio.

This project focuses on risk analysis of corporate bond portfolios. We divide the total risk of the portfolio into three parts, which are market risk, credit risk and liquidity risk. The market risk component is quantified by value-at-risk (VaR) which is determined by change in yield to maturity of the bond portfolio. For the credit risk component, we calculate default probabilities and losses in the event of default and then compute credit VaR. Next, we define a factor called `basis' which is the difference between the Credit Default Swap (CDS) spread and its corresponding corporate bond yield spread (z-spread or OAS). We quantify the liquidity risk by using the basis. In addition we also introduce a Fama-French multi-factor model to analyze the factor significance to the corporate bond portfolio.

Submitted by Maicon Juliano Schmidt (maicons) on 2015-04-08T12:07:56Z No. of bitstreams: 1 Samuel Ongaratto.pdf: 1337304 bytes, checksum: 14198512c9cbcad30e81483b17403288 (MD5) / Made available in DSpace on 2015-04-08T12:07:56Z (GMT). No. of bitstreams: 1 Samuel Ongaratto.pdf: 1337304 bytes, checksum: 14198512c9cbcad30e81483b17403288 (MD5) Previous issue date: 2010-08-20 / Nenhuma / O objetivo deste estudo é ajudar as empresas na tomada de decisão com relação à base de clientes e ajuste de suas promoções de mercado num contexto de risco de crédito comercial. O resultado da pesquisa é a proposição de um novo modelo. Como contribuições teóricas, pode ser citado o desenvolvimento de um modelo de risco de crédito capaz de estimar um risco de inadimplência para cada pagamento efetuado por um cliente. Isso difere dos modelos encontrados na literatura, que estimam apenas um risco para cada cliente. Outra contribuição é o desenvolvimento de um modelo baseado na métrica de Customer Lifetime Value com componentes inéditos (assimetria entre prazos de pagamento e recebimento e risco de crédito). Esta pesquisa é dividida em três fases distintas: uma fase exploratória, resultado de uma pesquisa realizada na literatura em busca de conceitos e elementos alinhados ao objetivo; a segunda é a proposição do método e do modelo propriamente ditos, e a terceira e última fase foi a aplicação do modelo a um estudo de campo, o qual utilizou dados de 14.259 faturas e 229 clientes. Os dados são de dezembro de 2007 a agosto de 2009. O modelo de risco de crédito integrado ao modelo proposto classifica as faturas pagas com 75,52% de assertividade média. Os resultados do estudo de campo ajudaram a empresa estudada a realizar uma série de mudanças na sua base de clientes, resultando com essas medidas num ganho estimado de mais de R$ 2,5 milhões para 2010. / The objective of this research is to help companies in decision making regarding the customer base and adjust their marketing promotions in the context of commercial credit risk. The result of this research is to propose a new model. As theoretical contributions can be mentioned the development of a model of credit risk can estimate a default risk for each payment made by a client. This differs from the other models in the literature that estimate only one risk for each client. Another contribution is the development of a model based on metrics of Customer Lifetime Value with components unpublished (asymmetry between receiving and payment terms and credit risk). This research is divided into three distinct phases: an exploratory phase, where it performed a literature review in search of concepts and elements aligned to the goal. The second phase is the proposition of the method and the model itself. The third and final phase was the implementation of the model to a field study. The field study used data from 14 259 bills and 229 customers. Data are from December 2007 to August 2009. The model of credit risk built into the proposed model classifies invoices paid on average 75.52% of assertiveness. The results of the field helped the company studied conducting a series of changes in its customer base. Changes made resulting in an estimated gain of more than R$ 2.5 million in 2010.

Submitted by Sara Ribeiro (sara.ribeiro@ucb.br) on 2018-08-08T13:33:03Z No. of bitstreams: 1 RafaelLeiteGregorioDissertacao2018.pdf: 1382550 bytes, checksum: 9c6e4f1d3c561482546aca581262b92b (MD5) / Approved for entry into archive by Sara Ribeiro (sara.ribeiro@ucb.br) on 2018-08-08T13:33:24Z (GMT) No. of bitstreams: 1 RafaelLeiteGregorioDissertacao2018.pdf: 1382550 bytes, checksum: 9c6e4f1d3c561482546aca581262b92b (MD5) / Made available in DSpace on 2018-08-08T13:33:24Z (GMT). No. of bitstreams: 1 RafaelLeiteGregorioDissertacao2018.pdf: 1382550 bytes, checksum: 9c6e4f1d3c561482546aca581262b92b (MD5) Previous issue date: 2018-07-09 / The credit risk assessment has a relevant role for financial institutions because it is associated with possible losses and has a large impact on the balance sheets. Although there are several researches on applications of machine learning and finance models, a study is still lacking that integrates available knowledge about credit risk assessment. This paper aims at specifying the machine learning model of the probability of default of publicly traded companies present in the Bovespa Index (corporations) and, based on the estimations of the model, to obtain risk assessment metrics based on risk letters. We converged methodologies verified in the literature and we estimated models that comprise fundamentalist (balance sheet) and governance data, macroeconomic and even variables resulting from the application of the proprietary model of KMV credit risk assessment. We test the XGboost and LinearSVM algorithms, which have very different characteristics among them, but are potentially useful to the problem. Parameter Grids were performed to identify the most representative variables and to specify the best performing model. The model selected was XGboost, and performance was very similar to the results obtained for the North American stock market in analogous research. The estimated credit ratings suggest that they are more sensitive to the economic and financial situation of the companies than that verified by traditional Rating Agencies. / A avaliação do risco de crédito tem papel relevante para as instituições financeiras por estar associada a possíveis perdas que podem gerar grande impacto nos balanços. Embora existam várias pesquisas sobre aplicações de modelos de aprendizado de máquina e finanças, ainda não há estudo que integre o conhecimento disponível sobre avaliação de risco de crédito. Este trabalho visa especificar modelo de aprendizado de máquina da probabilidade de descumprimento de empresas de capital aberto presentes no Índice Bovespa (corporações) e, fruto das estimações do modelo, obter métrica de avaliação de risco baseada em letras (ratings) de risco. Convergiu-se metodologias verificadas na literatura e estimou-se modelos que compreendem componentes fundamentalistas (de balanço) e de governança corporativa, macroeconômicos e ainda variáveis produto da aplicação do modelo proprietário de avaliação de risco de crédito KMV. Testou-se os algoritmos XGboost e LinearSVM, os quais possuem características bastante distintas entre si, mas são potencialmente úteis ao problema exposto. Foram realizados Grids de parâmetros para identificação das variáveis mais representativas e para a especificação do modelo com melhor desempenho. O modelo selecionado foi o XGboost, tendo sido observado desempenho bastante semelhante aos resultados obtidos para o mercado de ações norte-americano em pesquisa análoga. Os ratings de crédito estimados mostram-se mais sensíveis à situação econômico-financeira das empresas ante o verificado por agências de rating tradicionais.

Os modelos de análise e decisão de concessão de crédito buscam associar o perfil do tomador de crédito à probabilidade do não pagamento de obrigações contraídas, identificando assim o risco associado ao tomador e auxiliando a firma a decidir pela aprovação ou negação da solicitação de crédito. Atualmente este campo de pesquisa tem ganhado importância no cenário nacional - pela intensificação da atividade de crédito no país com grande participação dos bancos públicos neste processo - e internacional - pelo aumento das preocupações com potenciais danos à economia derivados de eventos de default. Tal quadro fez com que fossem construídos e adaptados diversos modelos e métodos à análise de risco de crédito tanto para consumidores como para empresas. Estes modelos são testados e comparados com base na acurácia de previsão ou de métricas de otimização estatística. Este é um procedimento que pode não se mostrar eficiente do ponto de vista financeiro, ao mesmo tempo em que dificulta a interpretação e tomada de decisão por parte da firma quanto a qual o melhor modelo, gerando uma lacuna pelo desprendimento observado entre a decisão de qual o modelo a ser adotado e o objetivo financeiro da empresa. Tendo em vista que o desempenho financeiro é um dos principais indicadores de qualquer procedimento gerencial, o presente estudo objetivou preencher a esta lacuna analisando o desempenho financeiro de carteiras de crédito formadas por técnicas de aprendizagem estatística utilizadas atualmente na classificação e análise de risco de crédito em pesquisas nacionais e internacionais. A pesquisa selecionou as técnicas: análise discriminante, regressão logística, redes bayesianas Naïve Bayes, kdB-1, kdB-2, SVC e SVM e aplicou tais técnicas junto à base de dados German Credit Data Set. Os resultados foram analisados e comparados inicialmente em termos de acurácia e custos por erro de classificação. Adicionalmente a pesquisa propôs o emprego de quatro métricas financeiras (RFC, PLR, RAROC e IS), encontrando variações quanto aos resultados produzidos por cada técnica. Estes resultados sugerem variações quanto a sequência de eficiência e consequentemente de emprego das técnicas, demonstrando a importância da consideração destas métricas para a análise e decisão de seleção de modelos de classificação ótimos. / Analysis and decision credit concession models search for relating the borrower\'s credit profile to the nonpayment probability of their obligations, identifying risks related to borrower and helping decision firm to approve or deny the credit request. Currently this search field has increased in Brazilian scenario - by credit activity intensification into the country with a large public banks sharing - and in the international scenario - by growing concerns about economy potential damages resulting from default events. This position leads the construction and adaptation of several models and methods by credit risk analysis from both consumers and companies. These models have been tested and compared based on prediction of accuracy or other statistical optimization metrics. This proceed is eventually not effective when analyzed by a financial standpoint, in the same time that affects the understanding and decision of the enterprise about the best model, creating a gap in the decision model choice and the firm financial goals. Given that the financial performance is a foremost indicator of any management procedure, this study aimed to address this gap by the financial performance analysis of loan portfolios formed by statistical learning techniques currently used in the classification and credit risk analysis in national and international researches. The selected techniques (discriminant analysis, logistic regression, Bayesian networks Naïve Bayes , 1 - KDB , KDB - 2 , SVC and SVM) were applied to the German Credit Data Set and their results were initially analyzed and compared in terms of accuracy and misclassification costs. Regardless of these metrics the research has proposed to use four financial metrics (RFC, PLR, RAROC and IS), finding variations in the results of each statistical learning techniques. These results suggest variations in the sequence of efficiency and, ultimately, techniques choice, demonstrating the importance of considering these metrics for analysis and selection of decision models of optimal classification.

This thesis focuses on analyzing the development of mortgage loans which were provided by Czech households. The first chapter is devoted to the theory of mortgage loans and their basic characteristics. In the second chapter I focus on credit risk. Its management, monitoring and reduction methods. The third chapter includes the causes of debt and the amount of elemental analysis based on macroeconomic variables of mortgage loans (mortgage loan rate, GDP, inflation, unemployment, gross wage). Proof of this dependence is part of the last chapter using regression and correlation analysis.

This thesis is about the credit default swap market and its development from the moment of its origin to the present. The focus is on the peripheral countries of eurozone, especially on Greece. The first part of the thesis is about the characteristics of CDS contracts, settlement of contracts and the relationship between CDS and insurance contracts. The other parts of the thesis are about the crisis in Greece, the CDS on sovereign greek bonds and the credit event. The last part of the thesis is about CDS on other sovereign bonds of peripheral countries in eurozone which are Spain, Italy and Portugal.