This master’s thesis project consists of two parts. Part I of the project compares modeling of historical resource production and forecasting of future production trends using the logit/probit transform advocated by Rutledge (2011) with conventional Hubbert curve fitting, using global coal production as a case study. The conventional Hubbert/Gaussian method fits a curve to historical production data whereas a logit/probit transform uses a linear fit to a subset of transformed production data. Within the errors and limitations inherent in this type of statistical modeling, these methods provide comparable results. That is, despite that apparent goodness-of-fit achievable using the Logit/Probit methodology, neither approach provides a significant advantage over the other in either explaining the observed data or in making future projections. For mature production regions, those that have already substantially passed peak production, results obtained by either method are closely comparable and reasonable, and estimates of ultimately recoverable resources obtained by either method are consistent with geologically estimated reserves. In contrast, for immature regions, estimates of ultimately recoverable resources generated by either of these alternative methods are unstable and thus, need to be used with caution. Although the logit/probit transform generates high quality-of-fit correspondence with historical production data, this approach provides no new information compared to conventional Gaussian or Hubbert-type models and may have the effect of masking the noise and/or instability in the data and the derived fits. In particular, production forecasts for immature or marginally mature production systems based on either method need to be regarded with considerable caution. Part II of the project investigates the utility of a novel alternative method for multicyclic Hubbert modeling tentatively termed “cycle-jumping” wherein overlap of multiple cycles is limited. The model is designed in a way that each cycle is described by the same three parameters as conventional multicyclic Hubbert model and every two cycles are connected with a transition width. Transition width indicates the shift from one cycle to the next and is described as weighted coaddition of neighboring two cycles. It is determined by three parameters: transition year, transition width, and γ parameter for weighting. The cycle-jumping method provides superior model compared to the conventional multicyclic Hubbert model and reflects historical production behavior more reasonably and practically, by better modeling of the effects of technological transitions and socioeconomic factors that affect historical resource production behavior by explicitly considering the form of the transitions between production cycles.

Esta dissertação analisa os impactos de fatores socioeconômicos e psicossociais sobre a prevalência da depressão no Brasil através de modelos probit e da decomposição de Oaxaca-Blinder aplicados às bases de dados dos suplementos de saúde da Pesquisa Nacional por Amostra de Domicílios (PNAD Saúde) e da Pesquisa Nacional de Saúde 2013 (PNS 2013). Neste trabalho, foram usados dois critérios básicos de identificação de indivíduos deprimidos: os que referiram ter recebido o diagnóstico de depressão de algum profissional de saúde e os que obtiveram um escore superior a 4 no teste PHQ-9 (cujas perguntas constam na PNS 2013). Foram obtidos resultados estatisticamente significativos, evidenciando a maior probabilidade da depressão entre mulheres e a relação inversa do transtorno com as variáveis renda domiciliar, desemprego, escolaridade e idade. Fatores como doenças físicas, doenças mentais e deficiências demonstraram uma relação direta com a depressão, embora as quantificações de suas magnitudes tenham sido sensíveis à especificação dos modelos. Também mostraram uma relação direta com o transtorno depressivo variáveis associadas a traumas e estresse emocional (como ter perdido um filho, ter sofrido algum tipo de violência, ter um filho com problemas de saúde, etc.). Outras variáveis como raça e região geográfica apresentaram resultados ambivalentes, também sensíveis às diferentes especificações de modelo. Ainda, variáveis referentes ao mercado de trabalho como o tipo de vínculo empregatício e setor de atividade apresentaram resultados inconclusivos. Por fim, outra conclusão importante foi que o critério de identificação dos indivíduos com depressão é determinante nos resultados. / This dissertation analyses the impacts of socioeconomic and psychosocial factors over the prevalence of depression in Brazil through probit models and the Oaxaca-Blinder decomposition applied to the data in Brazilian National Household Survey Health Supplement (PNAD Saúde) and the National Health Survey (PNS 2013). In this dissertation, two basic criteria were used to identify individuals with depression: those who declared to have received a depression diagnosis from a health professional and those who scored more than 4 in the PHQ-9 depression test (whose questions are included in PNS 2013). Statistically significant results were found, showing a higher probability of depression among women and an inverse relationship between the disorder and household income, unemployment, education and age. Factors such as chronic physical diseases, mental illnesses and deficiencies have showed a direct relationship with depression, although the quantification of such effects had a rather high sensitivity to model specification. Also, variables associated with emotional stress (such as having lost a child, being victim of some sort of violence, having a child with health problems) showed a direct relationship with depressive disorder. Other variables such as race and geographic region showed ambivalent results also very sensitive to different model specifications. Furthermore, labor market variables like type of work contract and activity sector show inconclusive results. Finally, another important finding is that the criteria for identifying individuals with depression were determinant for the results and conclusions.

abstract: Financing lease has bloomed as a new financing tool in China for the last several years. In this thesis I investigate the factors that influence China’s automobile financial leasing decisions by both lessors and lessees through market surveys. Based on Probit regression analysis of the data collected from 250 companies and 300 individuals, I find that a firm is more likely to use automobile financial leasing when its corporate tax rate is lower, growth potential is more stabilized, and profit is higher. It is also more likely to happen when a firm's long-term debt ratio and its degree of internationalization are higher. At the individual level, I find that the likelihood of individuals’ leasing decision is influenced by their risk preference, income level, and car price. Individuals’ gender, age and education level show no effect. Using the analytic hierarchy process (AHP) analysis, I further find that financing costs, service value-added, and products diversity are the three most important competitive factors for the auto financial leasing service providers. This is the case for both the corporate and individual customers in the sample. By contrast, the factors of sales channel and government relationship are found to be much less important. Finally, through an in-depth case study of the leasing company Shanghai Auto Financial Leasing, I find that the key factors determining the customers’ credit default risk are interest rate and automobile type. I also investigate factors that influence business risk during the automobile procurement stage, at the selling stage, and toward the disposition stage. The managerial implications of the above results are discussed throughout the thesis. / Dissertation/Thesis / Doctoral Dissertation Business Administration 2015

Made available in DSpace on 2014-06-12T15:52:11Z (GMT). No. of bitstreams: 2 arquivo3194_1.pdf: 1782444 bytes, checksum: 1982844f90df3787391d8faa431cde16 (MD5) license.txt: 1748 bytes, checksum: 8a4605be74aa9ea9d79846c1fba20a33 (MD5) Previous issue date: 2010 / Conselho Nacional de Desenvolvimento Científico e Tecnológico / Em redes neurais artificiais (RNAs), as funções de ativação mais comumente usadas são a função sigmóide logística e a função tangente hiperbólica, dependendo das características dos dados. Entretanto, a escolha da função de ativação pode influenciar fortemente o desempenho e a complexidade da rede neural. Neste trabalho, com o objetivo de melhorar o desempenho dos modelos de redes neurais, propomos o uso de novas funções de ativação no processamento das unidades da rede neural. Aqui, as funções não-lineares implementadas são as inversas das funções de ligação usadas em modelos de regressão binomial, essas funções são: complemento log-log, probit, log-log e Aranda, sendo que esta última função apresenta um parâmetro livre e é baseada na família de transformações Aranda-Ordaz. Uma avaliação dos resultados do poder de predição com estas novas funções através de simulação Monte Carlo é apresentada. Além disso, foram realizados diversos experimentos com aproximação de funções contínuas e arbitrárias, com regressão e com previsão de séries temporais. Na utilização da função de ativação com parâmetro livre, duas metodologias foram usadas para a escolha do parâmetro livre, l . A primeira foi baseada em um procedimento semelhante ao de busca em linha (line search). A segunda foi usada uma metodologia para a otimização global dessa família de funções de ativação com parâmetro livre e dos pesos das conexões entre as unidades de processamento da rede neural. A ideia central é otimizar simultaneamente os pesos e a função de ativação usada em uma rede multilayer perceptron (MLP), através de uma abordagem que combina as vantagens de simulated annealing, de tabu search e de um algoritmo de aprendizagem local. As redes utilizadas para realizar esses experimentos foram treinadas através dos seguintes algoritmos de aprendizagem: backpropagation (BP), backpropagation com momentum (BPM), backpropagation baseado no gradiente conjugado com atualizações Fletcher-Reeves (CGF) e Levenberg-Marquardt (LM)

Submitted by Israel Vieira Neto (israel.vieiraneto@ufpe.br) on 2015-03-04T14:13:21Z No. of bitstreams: 2 sammara_dissertacao.pdf: 1933363 bytes, checksum: 3948f7476054b8942a7b9c663ee83684 (MD5) license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) / Made available in DSpace on 2015-03-04T14:13:21Z (GMT). No. of bitstreams: 2 sammara_dissertacao.pdf: 1933363 bytes, checksum: 3948f7476054b8942a7b9c663ee83684 (MD5) license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) Previous issue date: 2012-08-10 / O primeiro ensaio propõe uma nova abordagem para estimar desigualdades socioeconômicas na Saúde Auto Avaliada. O método baseia-se em uma aplicação alternativa do Coeficiente de Gini, preservando a natureza categórica da variável e evitando incorrer nas dificuldades e limitações de cardinalizar tal indicador de saúde para análises de desigualdade. O proposto “Index-D” aplica as probabilidades preditas de um Modelo Probit Ordinal, sob a equação de Gini formatada para funções densidade discretas, permitindo-nos estimar variações na Saúde Auto Avaliada segundo subjacentes variações socioeconômicas e demográficas. Usamos dados da Pesquisa Nacional por Amostra de Domicílio (PNAD) para os anos de 1998, 2003 e 2008, apenas referente à população feminina, a fim de ilustrar a aplicabilidade do método. Os resultados mostram que a desigualdade em Saúde Auto Avaliada no Brasil decresceu de 1998 a 2008 entre as mulheres, independentemente do perfil socioeconômico considerado. Ainda, os grupos femininos com melhores condições financeiras apresentaram índices de desigualdade menores, enquanto que os mais pobres obtiveram os maiores escores. Considerando a desigualdade entre as Regiões, sudeste apresentou os resultados mais favoráveis, enquanto o Norte e o Nordeste reportaram as mais altas desigualdades, independentemente do ano e do perfil socioeconômico considerado. O segundo ensaio visa identificar o quanto da variância observada na Saúde Auto Avaliada (SAH) no Brasil é resultado do contexto onde as pessoas vivem. Dessa forma, através do Random-Intercept Ordered Probity Model, aplicamos uma amostra de municípios, retirada da Pesquisa Nacional por Amostra de Domicílios (PNAD), 2008, para representar as unidades do segundo nível, juntamente com as informações socioeconômicas dos indivíduos, a fim de controlar apropriadamente o efeito composição. Apesar de pequeno, o coeficiente de variação mostra a existência de variação sistemática na Saúde Auto Reportada entre os municípios urbanos do Brasil que persistiram mesmo após o controle do nível individual. As evidências sugerem que políticas de saúde no Brasil não devem investir apenas nas circunstâncias a nível individual, mas também sobre os ambientes sociais e físicos do coletivo, tais como segurança, espaços para lazer e infraestrutura urbana.

[pt] O mercado de capitais das grandes economias mundiais já incorporou o conceito de rating, ou seja, a classificação de risco de crédito, utilizada amplamente nos Estados Unidos há muitos anos. A existência de agências de rating capazes de fornecer classificações de risco de crédito totalmente independentes é uma condição imprescindível para o desenvolvimento de qualquer mercado de dívida. Este estudo tem como objetivo avaliar se as variáveis definidas no contrato de um bônus corporativo afetam o rating determinado por estas instituições para determinada emissão. Foram analisadas as variáveis cupom, preço de emissão, volume de emissão, maturidade e a presença de garantia. Além disso, analisou-se também a influência do risco país representado pelo EMBI +. Quanto ao método de análise, optou-se por comparar o método dos mínimos quadrados ordinários (MQO) com o probit ordenado. Os resultados mostram que não houve diferença de desempenho entre os dois modelos. E quanto às variáveis analisadas, apenas o cupom demonstrou exercer influência sobre o rating da emissão. / [en] The financial markets in the largest economies of the world often utilize rating agencies as a tool for credit risk classification, following the concept introduced in the USA a long time ago. The existence of independent institutions capable of classifying credit risk is a vital condition for the development of debt market. This study aims at analysing whether the variables defined in the indenture of the bond issue affect the rating assigned by rating agencies. The following variables were investigated: coupon, price, issued amount, maturity and the existence of collateral. Furthermore, it was analysed if the country risk, represented by the EMBI+ index, also affected the bond rating. Regarding the methodology, the ordinary least square (OLS) and ordered probit were chosen as the method of analysis. A comparative study was performed and the results indicated no major differences between both models. Concerning the variables analysed, only the coupon has shown some influence on the bond rating.

An important part when managing credit risk is to assess the probability of default of different counterparties. Increases and decreases in such probabil- ities are central components in the assessment, and this is where transition matrices become useful. These matrices are commonly used tools when as- sessing counterparty credit risk, and contain the probability of default, as well as the probability to migrate between different predefined rating classifica- tions. These rating classifications are used to reflect the risk taken towards different counterparties. Therefore, it is important for financial institutions to develop accurate transition matrix models to manage predicted changes in credit risk exposure. This is because counterparty creditworthiness and prob- ability of default indirectly affect expected loss and the capital requirement of held capital. This thesis will analyze how two specific models perform when used for generating transition matrices. These models will be tested to investigate their performance when predicting rating transitions, including probability of default. / En viktig del vid hanteringen av kreditrisk är att bedöma sannolikheten för fallissemang för olika motparter. Ökningar och minskningar i dessa sanno- likheter är centrala komponenter i bedömningen, och det är här migrations- matriser blir användbara. Dessa matriser är vanligt förekommande verktyg vid bedömning av kreditrisk mot olika motparter och innehåller sannolikheten för fallissemang samt sannolikheten att migrera mellan olika fördefinierade be- tygsklassificeringar. Dessa betygsklassificeringar används för att återspegla den risk som tas mot olika motparter. Det är därför viktigt för finansinstitut att utveckla träffsäkra migrationsmatris modeller för att hantera förväntade förändringar i kreditriskexponering. Detta beror på att kreditvärdigheten hos motparter samt sannolikheten för fallissemang indirekt påverkar expected loss och kapitalkrav. Detta examensarbete kommer att analysera hur två specifika modeller presterar när de används för att generera migrationsmatriser. Dessa mod- eller kommer att testas för att undersöka hur de presterar när de används för att förutsäga övergångar inom betygsklassificering, inklusive sannolikheten för fallissemang.

Correlated ordinal data typically arise from multiple measurements on a collection of subjects. Motivated by an application in credit risk, where multiple credit rating agencies assess the creditworthiness of a firm on an ordinal scale, we consider multivariate ordinal models with a latent variable specification and correlated error terms. Two different link functions are employed, by assuming a multivariate normal and a multivariate logistic distribution for the latent variables underlying the ordinal outcomes. Composite likelihood methods, more specifically the pairwise and tripletwise likelihood approach, are applied for estimating the model parameters. We investigate how sensitive the pairwise likelihood estimates are to the number of subjects and to the presence of observations missing completely at random, and find that these estimates are robust for both link functions and reasonable sample size. The empirical application consists of an analysis of corporate credit ratings from the big three credit rating agencies (Standard & Poor's, Moody's and Fitch). Firm-level and stock price data for publicly traded US companies as well as an incomplete panel of issuer credit ratings are collected and analyzed to illustrate the proposed framework. / Series: Research Report Series / Department of Statistics and Mathematics

Les TIC sont un facteur clé de performance dans les pays développés. Cette thèse s’intéresse à l’adoption des TIC et leur impact sur la performance des PME manufacturières d’un pays en développement. A la suite d’une première partie qui présente le cadre théorique et conceptuel, le reste de la thèse est organisé en trois études empiriques. La première étude propose une modélisation Probit afin d’identifier les déterminants d’adoption des TIC. Le capital humain est la variable explicative la plus significative. Se basant sur la régression linéaire à variables muettes, la causalité de Granger, le test de Kruskal-Wallis et le test de l’ANOVA de Welch, suivis des tests post-hoc correspondants, la deuxième étude met en évidence l’existence d’un fort lien statistique significatif entre le niveau d’adoption des TIC et la rentabilité. Dans une troisième étude, plusieurs modélisations Probit (simple, ordonné et multivarié) ont été testées sur différentes mesures de performance. Nous montrons, premièrement, que les TIC ont un impact positif sur la productivité, la rentabilité et la compétitivité. Deuxièmement, les TIC, le capital humain et la formation sont les déterminants de la performance globale. Enfin, la contribution des TIC à la performance globale est forte lorsqu’elles sont combinées au capital humain qualifié. En définitive, nos résultats empiriques ont montré un effet positif des TIC, du capital humain et du changement organisationnel sur la performance des PME. / ICT is a key performance factor in developed countries. This PhD thesis focuses on the adoption of ICTs and their impact on the performance of manufacturing SMEs in a developing country. Following a first part covering the theoretical and conceptual framework, the rest of the thesis is organized in three empirical studies. The first study uses a Probit model in order to identify the determinants of ICT adoption. Human capital seems to be the most significant explanatory variable. Based on linear regression of dummy variables, Granger causality, Kruskal-Wallis test, ANOVA test of Welch, followed by corresponding post-hoc tests, the second study highlights the existence of a strong statistically significant relationship between the level of ICT adoption and profitability. In a third study, many Probit models (simple, ordered and multivariate) were tested on different measures of performance. Firstly, we show that ICT have a positive impact on productivity, profitability and competitiveness of SMEs. Secondly, ICT, human capital and training are determinants of firm overall performance. Thirdly, when combined together, ICT and highly skilled human resources have an important contribution to the global performance. In conclusion, our empirical results demonstrate a positive impact of ICT, human capital and organizational change on firm performance.

Credit Rating Agencies (CRAs) play a key role in financial markets by helping to reduce informative asymmetry between lenders and investors, on one side, and issuers on the other side, with regard to the creditworthiness of banks or countries. This crucial role has expanded alongside financial globalisation and received an additional boost from Basel II which integrates the ratings of CRAs into the rules for setting weights for credit risk. Ratings adjustment tends to be sticky, lagging behind markets, and often overreact when they do change. This overreaction may have aggravated the recent financial crises, contributing to financial instability and cross-country contagion. Criticism has been especially directed towards the high degree of concentration of the ratings industry. Promotion of competition may require policy action at the international level to encourage the establishment of new agencies and to discover alternative rules or regulatory requirements in order to achieve promising results. The recent growth of Middle Eastern and North African countries (MENA) and their commercial banking system has increased the need of paying widespread attention to this region of the world. This thesis crucially identifies, and estimates, the robust determinants of credit ratings for MENA countries and their commercial banks, incorporating a set of bank level accounting and financial risk factors, as well as country-specific characteristics, including indicators for regulatory, supervision, legal and economic environments. The research contributes, firstly, to the theoretical literature on credit ratings industry by reviewing extant methodologies specifically as they apply to banks and sovereign countries. Secondly, it conducts a systematic, cross-country empirical investigation using panel data econometric methodology for the purpose of estimating MENA countries sovereign and bank credit rating models. Thirdly, it provides tangible and statistically significant evidence on the different factors that determines the estimation of credit ratings and influencing bank's risk. The extant literature reviewed serves as a basis to achieve and develop the research aim, objectives and hypotheses of the thesis. The research then constructs an appropriate panel dataset from different sources, containing bank-level and country-level information for a sample of 108 commercial banks covering 13 MENA countries over the period 2000 - 2012. The methodological framework for estimating credit rating models (linear regression, logit and probit) is also reviewed and the procedures for panel data estimation are implemented using the econometric package STATA (version 13). All relevant data are drawn from public sources including Reuters, Bankscope, IMF and the World Bank. Using the random effects ordered probit and logit methodologies to estimate both sovereign (country) and bank level credit ratings models for the MENA countries, the evidence shows that real GDP growth, capital requirements, restrictions on banking activities and control of corruption all contribute negatively to the sovereign ratings. Furthermore, internal management and organisational requirements is considered as an additional regulatory factor not studied in previous research. The statistically significant and inverse relationship of the latter is considered an important and interesting outcome of MENA countries’ sovereign ratings. On the other hand, GDP per capita, investment (as a percentage of GDP), political stability, government effectiveness and the rule of law all reveal significant and positive impact on the sovereign credit ratings. In general, this research finds that improved macroeconomic conditions are correlated with higher ratings, while greater reserve regulations are correlated with lower ratings. The study also does find the significance of governance and regulatory variables plays a key role into the final credit rating. With regard to the impact on banks’ ratings, the results show that higher return on average assets and equity, larger bank size, more restrictions on bank activities, as well as higher official disciplinary power and higher standards of internal management, will yield higher credit ratings. Apart from having direct and positive impact on banks credit ratings, these variables are important for examining the risk-sharing incentives in MENA countries’ banks. In contrast, the estimation results indicate that net interest margin, net loans to deposits, liquid assets to deposits, capital requirements, deposit insurance scheme, liquidity requirements, unemployment rate and government effectiveness have an inverse and negative impact on banks ratings. In general, this study also finds various financial, macroeconomic, and regulatory effects on banks’ credit ratings. To a much lesser extent than government ratings, various macroeconomic variables also helped predict banks’ ratings, including real GDP growth and the unemployment rate. The thesis concludes by arguing that the combined use of financial and non-financial factors for estimating credit ratings models supports the relevant hypotheses examined and adds value to all stakeholders in improving and obtaining a better quality of credit ratings. This study also demonstrates that a diversity of bank-level and country-level factors influence the MENA sovereign and bank ratings differently, implying that policy makers, regulators alongside rating agencies should distinguish the different environmental factors between nations before any judgment and issuance can be model of the ratings. To conclude, there is no study which exclusively investigates credit rating models for the MENA region exploiting the richness of the data and methodology employed, and the current research aims to fill this gap.