La complexité des marchés financiers et la recrudescence des crises particulièrement sévères contribuent à l'évolution et à la remise en cause de modèles économétriques dits standards dans l'explication et la prévision des dynamiques financières. L'alerte donnée conjointement par les responsables prudentiels et les chercheurs vise à encourager le développement de modèles plus complexes, non linéaires et largement inspirés d'autres disciplines. Nous soutenons dans cette thèse l'idée qu'une approche chaos-stochastique des chroniques financières est susceptible de conduire à de meilleurs résultats. La pertinence de cette association est évaluée pour le risque de marché dans deux cadres d'analyse distincts. Nous montrons tout l'intérêt d'une synthèse des modèles chaotiques et des spécifications GARCH avec ou sans changements de régimes markoviens (MRS) pour la modélisation et la prévision de la Value-at-Risk des indices boursiers de la zone euro. Il ressort de cette étude de meilleurs résultats des modèles chaos-stochastiques et dans le cas des spécifications MRS-GARCH, une meilleure adéquation du modèle chaotique de Lasota(1977) pour les indices de l'Europe du Sud, particulièrement plus volatiles que ceux de l'Europe du Nord pour lesquels nous recommandons le modèle de Mackey-Glass(1977). Cette combinaison permet, dans un cadre bivarié, de mieux appréhender les liens qui existent entre les différentes places boursières de la zone euro. Nous introduisons deux nouvelles spécifications qui intègrent les problématiques liées aux ruptures de corrélations : la première permet de distinguer, par une analyse en sous-périodes, les relations d'interdépendance par rapport aux phénomènes de contagion et la seconde propose, dans un cadre unifié, d'intégrer les ruptures de corrélations. Cette double analyse met en évidence le rôle moteur du couple d'indices franco-allemand, l'existence de deux sphères distinctes constituées d'une part par les indices de l'Europe du Nord et d'autre part par les pays de l'Europe du Sud et l'intensification de certaines relations entre indices suite à la crise des dettes souveraines. Nous constatons et insistons sur la pertinence d'un modèle chaotique en moyenne pour rendre compte d'une part de la volatilité attribuée, à tort, aux effets GARCH. / The complexity of financial markets and the resurgence of severe crises contribute to the skepticism and evolution of standard econometric models in the explanation and prediction of financial time series. The warning given jointly by prudential authorities and researchers aims to encourage the development of nonlinear and more complex models inspired by other disciplines. I argue in this thesis that a chaos-stochastic approach of financial dynamics is likely to lead to better results. The relevance of this association is evaluated for market risk in two distinct analytical frameworks. I show the improvements given by a synthesis of chaotic models and GARCH specifications with or without Markov Regime Switching (MRS) for modelling and predicting the Value-at-Risk of 7 mains index of Monetary and Economic Union. It appears, from this study, better results from chaos-stochastic models. In the case of the MRS-GARCH specifications, I find more adequacy of the chaotic model of Lasota (1977) for the indices of Southern Europe, which are especially more volatile than those of Northern Europe for which I recommend the model of Mackey-Glass (1977). This combination allows, in a bivariate framework, to provide information on the relationship between these different indices. I introduce two new specifications that integrate issues related to correlation breakdowns. The first distinguishes, by a sub-periods analysis, the relations of interdependence of contagious relationships. Meanwhile, the second provides, in a unified framework, an integration of correlations breakdowns. These two analyses imply It appears from this double analysis the leading role of the Franco-German duo, the existence of two distinct spheres formed in a part by the Northern European indices and in another part by countries of the Southern Europe, and the intensification of relations between some indices following the sovereign debt crisis. Finally, these results support the relevance of a chaotic model which may account for some volatilities that are, wrongly, attributed to GARCH effects.

Durante os últimos anos, tem havido muitas mudanças na maneira como as instituições financeiras avaliam o risco. As regulações têm tido um papel muito importante no desenvolvimento das técnicas de medição do risco. Diante das diversidades de técnicas de estimação e análise de risco utilizadas pelas bolsas de valores e de futuros, nacionais e internacionais, bem como as Clearings de controle de risco, este estudo propôs uma análise comparativo de modelos de volatilidade para o cálculo do Value-at-Risk (VaR) aplicados aos principais índices de ações do mercado financeiro internacional. Utilizouse os modelos de volatilidade condicional da família ARCH levando em consideração a presença de longa dependência em seus retornos (memória longa) e assimetria na volatilidade. Em específico, utilizaram-se os modelos GARCH, EGARCH, APARCH, FIGARCH, FIEGARCH, FIAPARCH e HYGARCH estimados a parir de quatro diferentes distribuições, Normal, t-Student, G.E.D. e t-Student Assimétrica. Analisaramse os índices dos principais mercados de ações do mundo, sendo: Dow Jones, S&P 500, Nasdaq, Ibovespa, FTSE e Nikkei 225. Testou-se também a capacidade preditiva do modelo Riskmetrics desenvolvido pelo J.P. Morgan para o calculo do VaR, comparado com os modelos de volatilidade. Os resultados obtidos sugerem que o pacote desenvolvido pelo J.P.Morgan não se aplica adequadamente à realidade do mercado acionário mundial, como ferramenta de gestão e controle do risco das oscilações dos preços das ações de empresas negociadas nas bolsas de Nova Iorque, Nasdaq, BM&FBOVESPA, bolsa de Londres e bolsa de Tóquio. Os modelos que consideram o efeito de memória longa na volatilidade condicional dos retornos dos índices, em especial o modelo FIAPARCH (1,d,1), foram os que obtiveram melhor ajuste e desempenho preditivo do risco de mercado (Value-at-Risk), conforme valores apresentados pelo teste de razão de falha proposto por Kupiec (1995). / In recent years, there have been many changes in how financial institutions assess risk. The regulations have had a very important role in the development of techniques for measuring risk. Considering the diversity of estimation techniques and risk analysis used by stock exchanges and futures, national and international, as well as clearing houses of risk control, this study proposed a comparative analysis of volatility models for calculating Value-at-Risk (VaR) to the major stock indexes of international finance. It used models of conditional volatility of the ARCH family taking into account the presence of long dependence on their returns (long memory) and asymmetry in volatility. Specifically, it used the models GARCH, EGARCH, APARCH, FIGARCH, FIEGARCH, FIAPARCH and HYGARCH estimated the birth of four different distributions, Normal, t-Student, GED and t-Student Asymmetric. It analyzed the contents of the major stock markets of the world, being: Dow Jones, S & P 500, NASDAQ, Bovespa index, FTSE and Nikkei 225. Was also tested the predictive ability of the RiskMetrics model developed by JP Morgan for the calculation of VaR, compared with the models of volatility. The results suggest that the package developed by JPMorgan does not apply adequately to the reality of global stock market as a tool to manage and control the risk of fluctuations in stock prices of companies traded on the New York Stock Exchange, Nasdaq, BM&FBOVESPA, London Stock Exchange and Tokyo Stock Exchange. Models that consider the effect of long memory in conditional volatility of returns of the indices, especially the model FIAPARCH (1, d, 1), were the ones showing better fit and predictive performance of market risk (Value-at-Risk) , according to figures provided by the ratio test proposed by Kupiec (1995).

No description available.

No description available.

Malgré la multiplicité des méthodes d'estimation de la VaR, elles souffrent d'une faiblesse fondamentale. En effet, elles ne font aucune distinction entre l'information captée à basse fréquence et celle captée à haute fréquence. Ce qui revient à  supposer de façon implicite que l'information contenue dans les données historiques a la même importance quel que soit l'horizon temporel de l'investisseur c'est-à-dire sa fréquence de trading (intra-journalière, journalière, hebdomadaire, mensuelle,..). Mais, accepter une telle hypothèse revient à supposer que les marchés financiers sont homogènes. Ce qui est contraire à la réalité empirique. En effet, les marchés financiers sont caractérisés par une grande hétérogénéité d'acteurs. L'objet de notre thèse est d'apporter une contribution à l'estimation de la VaR basée sur la décomposition de la volatilité dans le domaine des fréquences. Ce qui nous permet de mette en évidence l'influence de l'hétérogénéité des horizons temporels des acteurs des marchés financiers sur l'estimation de la Value at Risk. Pour cela,nous faisons appel à un outil statistique susceptible de nous procurer de l'information temporelle sur la volatilité et de l'information fréquentielle sur la fréquence de trading des différents acteurs des marchés financiers: l'approche temps-fréquence de la transformée en ondelettes. / Although multiplicity of VaR estimate approaches,they suffer from a fundamental weakness.They don't make any distiction between informations captured in a high frequency and in a low frequency manner.It is an implicit assumption of homogeneity of fiancial markets in contrast to empirical facts. In our thesis, we try to construct a VaR model based on volatility decomposition in the frequency domain.It enables us to show how the time horizons heterogeneity of financial markets participants could influence value at risk estimates.We use a statistical tool able to give us temporal information about volatility and frequencial information about trading frequencies of market participants:the time frequency approach of wavelet transform.

A avaliação de riscos em projetos de produtos inovadores transformou-se em peça chave, na atualidade de competição crescente, para as empresas. Esse fato foi reconhecido na área de pesquisa nas últimas décadas e vários autores desenvolveram modelos para estimar riscos em projetos de produtos inovadores, tanto qualitativos como quantitativos. Porém, não foram encontrados, nas principais bases de dados, estudos que tratam o tema pela integração dos riscos de desenvolvimento e comercialização, mensurados por meio do Value at Risk (VaR). O objetivo geral do trabalho, portanto, é propor um modelo de análise de investimento que integre as etapas de desenvolvimento e comercialização para projetos inovadores, com utilização do VaR como medida de risco. Para a elaboração do modelo foi desenvolvido, primeiramente, um framework, qual relaciona os principais riscos em projetos de inovação e as variáveis que quantificam as tais. Este framework serve como base para a construção do modelo conceitual. Com a utilização das variáveis no modelo é possível estimar e quantificar os processos de desenvolvimento. A aplicação do VaR para a avaliação econômica em projetos de investimento representa uma novidade, mas se baseia na aplicação normal de estimação de riscos desenvolvida para o mercado financeiro. A vantagem do VaR é que resume os riscos considerados no cálculo do projeto em um único número, em unidades monetárias e de fácil compreensão, o que permite a comparação de projetos de investimento mutuamente exclusivos. O modelo integrado proposto possibilita uma avaliação econômica mais tangível que os métodos tradicionais de avaliação, aproximando o resultado à realidade e assim apresentando um avanço na estimação de risco no ambiente de desenvolvimento de produtos inovadores. Este fato foi comprovado na aplicação do método proposto em duas simulações de casos reais, quais resultados foram consistentes e compreensíveis. / Risk assessment in innovative product projects has become a key point for companies in today\'s growing competition. This fact was recognized by research in the area in recent decades and several authors have developed models to estimate risks in innovative product projects, as well as qualitative and quantitative. However, in the main databases could not be found studies dealing with the issue by integrating the risks of the development phase and the commercialization phase, measured by Value at Risk (VaR). The overall objective of this work is, therefore, proposing an investment analysis model that integrates the stages of development and commercialization for innovative projects, using VaR as a risk measure. Firstly, a framework, which relates the main risks in innovation projects and the variables that quantify such, was developed. This framework serves as a basis for the construction of the conceptual model. With the use of the defined variables and the conceptual model it is possible to estimate and quantify the processes of innovation projects. The application of VaR for economic evaluation of investment projects is new, but it is based on the risk estimates application used in the financial market. The advantage of VaR methods is that they summarize the risks considered in the project calculation in a single number expressed in monetary units, which is easy to interpret, allowing the comparison of mutually exclusive investment projects. The proposed integrated model enables a more tangible economic assessment than traditional methods of evaluation, bringing the result closer to reality and thus presenting an advance in risk estimation in innovative product development environment. This was proven in the application of the proposed method in two simulations of real cases, which results were consistent and understandable.

Diante das exigências estipuladas pelos órgãos reguladores pelos acordos internacionais, tendo em vistas as inúmeras crises financeiras ocorridas nos últimos séculos, as instituições financeiras desenvolveram diversas ferramentas para a mensuração e controle do risco inerente aos negócios. Apesar da crescente evolução das metodologias de cálculo e mensuração do risco, o Value at Risk (VaR) se tornou referência como ferramenta de estimação do risco de mercado. Nos últimos anos novas técnicas de cálculo do Value at Risk (VaR) vêm sendo desenvolvidas. Porém, nenhuma tem sido considerada como a que melhor ajusta os riscos para diversos mercados e em diferentes momentos. Não existe na literatura um modelo conciso e coerente com as diversidades dos mercados. Assim, o presente trabalho tem por objetivo geral avaliar os estimadores de risco de mercado, gerados pela aplicação de modelos baseados no Value at Risk (VaR), aplicados aos índices das principais bolsas dos países desenvolvidos e emergentes, para os períodos normais e de crise financeira, de modo a apurar os mais efetivos nessa função. Foram considerados no estudo os modelos VaR Não condicional, pelos modelos tradicionais (Simulação Histórica, Delta-Normal e t-Student) e baseados na Teoria de Valores Extremos; o VaR Condicional, comparando os modelos da família ARCH e Riskmetrics e o VaR Multivariado, com os modelos GARCH bivariados (Vech, Bekk e CCC), funções cópulas (t-Student, Clayton, Frank e Gumbel) e por Redes Neurais Artificiais. A base de dados utilizada refere-se as amostras diárias dos retornos dos principais índices de ações dos países desenvolvidos (Alemanha, Estados Unidos, França, Reino Unido e Japão) e emergentes (Brasil, Rússia, Índia, China e África do Sul), no período de 1995 a 2013, contemplando as crises de 1997 e 2008. Os resultados do estudo foram, de certa forma, distintos das premissas iniciais estabelecidas pelas hipóteses de pesquisa. Diante de mais de mil modelagens realizadas, os modelos condicionais foram superiores aos não condicionais, na maioria dos casos. Em específico o modelo GARCH (1,1), tradicional na literatura, teve uma efetividade de ajuste de 93% dos casos. Para a análise Multivariada, não foi possível definir um modelo mais assertivo. Os modelos Vech, Bekk e Cópula - Clayton tiveram desempenho semelhantes, com bons ajustes em 100% dos testes. Diferentemente do que era esperado, não foi possível perceber diferenças significativas entre os ajustes para países desenvolvidos e emergentes e os momentos de crise e normal. O estudo contribuiu na percepção de que os modelos utilizados pelas instituições financeiras não são os que apresentam melhores resultados na estimação dos riscos de mercado, mesmo sendo recomendados pelas instituições renomadas. Cabe uma análise mais profunda sobre o desempenho dos estimadores de riscos, utilizando simulações com as carteiras de cada instituição financeira. / Given the requirements stipulated by regulatory agencies for international agreements, in considering the numerous financial crises in the last centuries, financial institutions have developed several tools to measure and control the risk of the business. Despite the growing evolution of the methodologies of calculation and measurement of Value at Risk (VaR) has become a reference tool as estimate market risk. In recent years new calculation techniques of Value at Risk (VaR) have been developed. However, none has been considered the one that best fits the risks for different markets and in different times. There is no literature in a concise and coherent model with the diversity of markets. Thus, this work has the objective to assess the market risk estimates generated by the application of models based on Value at Risk (VaR), applied to the indices of the major stock exchanges in developed and emerging countries, for normal and crisis periods financial, in order to ascertain the most effective in that role. Were considered in the study models conditional VaR, the conventional models (Historical Simulation, Delta-Normal and Student t test) and based on Extreme Value Theory; Conditional VaR by comparing the models of ARCH family and RiskMetrics and the Multivariate VaR, with bivariate GARCH (VECH, Bekk and CCC), copula functions (Student t, Clayton, Frank and Gumbel) and Artificial Neural Networks. The database used refers to the daily samples of the returns of major stock indexes of developed countries (Germany, USA, France, UK and Japan) and emerging (Brazil, Russia, India, China and South Africa) from 1995 to 2013, covering the crisis in 1997 and 2008. The results were somewhat different from the initial premises established by the research hypotheses. Before more than 1 mil modeling performed, the conditional models were superior to non-contingent, in the majority of cases. In particular the GARCH (1,1) model, traditional literature, had a 93% adjustment effectiveness of cases. For multivariate analysis, it was not possible to set a more assertive style. VECH models, and Bekk, Copula - Clayton had similar performance with good fits to 100% of the tests. Unlike what was expected, it was not possible to see significant differences between the settings for developed and emerging countries and the moments of crisis and normal. The study contributed to the perception that the models used by financial institutions are not the best performing in the estimation of market risk, even if recommended by renowned institutions. It is a deeper analysis on the performance of the estimators of risk, using simulations with the portfolios of each financial institution.

Com a crescente demanda de modelagem de riscos dependentes, enfatizamos neste trabalho a teoria de cópulas e algumas medidas de dependência tais como coeficiente de correlação linear, coeficiente de correlação de Spearman. Mostramos algumas interpretações errôneas sobre o coeficiente de correlação linear e como podemos realizar simulações de variáveis aleatórias com determinadas marginais e dependência. Realizamos uma aplicação na área de seguros para determinar o capital alocado da seguradora. / With the growing demand for modeling dependent risk, in this study we emphasize the theory of copulas and some measures of dependence such as linear correlation coefficient and Spearman correlation coefficient. We show some misleading interpretations on the linear correlation coefficient, and how we can perform simulations of random variables with some marginals and dependence. We conduct an application in the insurance area to determine the allocated capital of the insurer.

Cette thèse se concentre sur des mesures du risque financier et la modélisation de la volatilité. L’objectifgénéral est : (i) de proposer de nouvelles techniques pour mesurer à la fois le risque systémique et lerisque à haute fréquence, et (ii) d’appliquer et d’améliorer les outils économétriques de modélisation etde prévision de la volatilité. Ce travail comporte quatre chapitres (papiers de recherche).La première partie de la thèse traite des questions liées à la modélisation et la prévision des mesuresdu risque à haute fréquence et du risque systémique. Plus précisément, le deuxième chapitre proposeune nouvelle mesure du risque systémique utilisée pour identifier les institutions financières d’importancesystémique (SIFIs). Basée sur une approche spécifique, cette mesure originale permet de décomposer lerisque global du système financier tout en tenant compte des caractéristiques de l’entreprise. Le troisièmechapitre propose des mesures du risque de marché intra-journalier dans le contexte particulier des donnéesà haute fréquence irrégulièrement espacées dans le temps (tick-by-tick).La deuxième partie de la thèse est consacrée aux méthodes d’estimation et de prévision de la volatilitéincluant directement des données à haute fréquence ou des mesures réalisées de volatilité. Ainsi, dans lequatrième chapitre, nous cherchons à déterminer, dans le contexte des modèles de mélange des fréquencesd’échantillonnage (MIDAS), si des regresseurs à haute fréquence améliorent les prévisions de la volatilitéà basse fréquence. Une question liée est de savoir s’il existe une fréquence d’échantillonnage optimaleen termes de prévision, et non de mesure de la volatilité. Le cinquième chapitre propose une versionrobuste aux jumps du modèle Realized GARCH. L’application porte sur la crise / This thesis focuses on financial risk measures and volatility modeling. The broad goal of this dissertationis: (i) to propose new techniques to measure both systemic risk and high-frequency risk, and (ii) toapply and improve advanced econometric tools to model and forecast time-varying volatility. This workhas been concretized in four chapters (articles).The first part addresses issues related to econometric modeling and forecasting procedures on bothsystemic risk and high-frequency risk measures. More precisely, Chapter 2 proposes a new systemic riskmeasure used to identify systemically important financial institutions (SIFIs). Based on a componentapproach, this original measure allows to decompose the risk of the aggregate financial system whileaccounting for the firm characteristics. Chapter 3 studies the importance and certifies the validity ofintraday High Frequency Risk (HFR) measures for market risk in the special context of irregularly spacedhigh-frequency data.The second part of this thesis tackles the need to improve the estimation/prediction of volatility bydirectly including high-frequency data or realized measures of volatility. Therefore, in Chapter 4 weexamine whether high-frequency data improve the volatility forecasts accuracy, and if so, whether thereexists an optimal sampling frequency in terms of prediction. Chapter 5 studies the financial volatilityduring the global financial crisis. To this aim, we use the largest volatility shocks, as provided by therobust version of the Realized GARCH model, to identify and analyze the events having induced theseshocks during the crisis.

Com a crescente demanda de modelagem de riscos dependentes, enfatizamos neste trabalho a teoria de cópulas e algumas medidas de dependência tais como coeficiente de correlação linear, coeficiente de correlação de Spearman. Mostramos algumas interpretações errôneas sobre o coeficiente de correlação linear e como podemos realizar simulações de variáveis aleatórias com determinadas marginais e dependência. Realizamos uma aplicação na área de seguros para determinar o capital alocado da seguradora. / With the growing demand for modeling dependent risk, in this study we emphasize the theory of copulas and some measures of dependence such as linear correlation coefficient and Spearman correlation coefficient. We show some misleading interpretations on the linear correlation coefficient, and how we can perform simulations of random variables with some marginals and dependence. We conduct an application in the insurance area to determine the allocated capital of the insurer.