Cette thèse traite de l’optimisation multiobjectif de fonctions coûteuses, aboutissant à laconstruction d’un front de Pareto représentant l’ensemble des compromis optimaux. En conception automobile, le budget d’évaluations est fortement limité par les temps de simulation numérique des phénomènes physiques considérés. Dans ce contexte, il est courant d’avoir recours à des « métamodèles » (ou modèles de modèles) des simulateurs numériques, en se basant notamment sur des processus gaussiens. Ils permettent d’ajouter séquentiellement des observations en conciliant recherche locale et exploration. En complément des critères d’optimisation existants tels que des versions multiobjectifs du critère d’amélioration espérée, nous proposons d’estimer la position de l’ensemble du front de Pareto avec une quantification de l’incertitude associée, à partir de simulations conditionnelles de processus gaussiens. Une deuxième contribution reprend ce problème à partir de copules. Pour pouvoir traiter le cas d’un grand nombre de variables d’entrées, nous nous basons sur l’algorithme REMBO. Par un tirage aléatoire directionnel, défini par une matrice, il permet de trouver un optimum rapidement lorsque seules quelques variables sont réellement influentes (mais inconnues). Plusieurs améliorations sont proposées, elles comprennent un noyau de covariance dédié, une sélection du domaine de petite dimension et des directions aléatoires mais aussi l’extension au casmultiobjectif. Enfin, un cas d’application industriel en crash a permis d’obtenir des gainssignificatifs en performance et en nombre de calculs requis, ainsi que de tester le package R GPareto développé dans le cadre de cette thèse. / This dissertation deals with optimizing expensive or time-consuming black-box functionsto obtain the set of all optimal compromise solutions, i.e. the Pareto front. In automotivedesign, the evaluation budget is severely limited by numerical simulation times of the considered physical phenomena. In this context, it is common to resort to “metamodels” (models of models) of the numerical simulators, especially using Gaussian processes. They enable adding sequentially new observations while balancing local search and exploration. Complementing existing multi-objective Expected Improvement criteria, we propose to estimate the position of the whole Pareto front along with a quantification of the associated uncertainty, from conditional simulations of Gaussian processes. A second contribution addresses this problem from a different angle, using copulas to model the multi-variate cumulative distribution function. To cope with a possibly high number of variables, we adopt the REMBO algorithm. From a randomly selected direction, defined by a matrix, it allows a fast optimization when only a few number of variables are actually influential, but unknown. Several improvements are proposed, such as a dedicated covariance kernel, a selection procedure for the low dimensional domain and of the random directions, as well as an extension to the multi-objective setup. Finally, an industrial application in car crash-worthiness demonstrates significant benefits in terms of performance and number of simulations required. It has also been used to test the R package GPareto developed during this thesis.

[pt] O objetivo desta tese é apresentar novas alternativas para modelagem de variáveis aleatórias no setor de seguros, utilizando o arcabouço dos modelos orientados por score com parâmetros variantes no tempo. No primeiro artigo, propomos um modelo dinâmico para a distribuição do sinistro agregado, que corresponde à soma aleatória dos valores de sinistros (severidade) em determinado período de tempo. A obtenção da distribuição do sinistro agregado é um problema clássico na teoria do risco e fundamental para precificação de seguros, cálculo de provisões e de probabilidade de ruína. No entanto, a obtenção da expressão analítica para essa distribuição de probabilidade é uma tarefa difícil. Neste trabalho, especificamos distribuições não-Gaussianas, tanto para o número de sinistros como para severidade, sob o arcabouço GAS (Generalized Autoregressive Score), e, por meio do uso da Transformada Rápida de Fourier obtemos, numericamente, a distribuição do sinistro agregado. O segundo artigo trata da incorporação do efeito de variáveis macroeconômicas na modelagem de variáveis relevantes no setor de seguros, em linha com requisito internacional de avaliação de provisões de forma consistente com mercado, a qual leva em consideração as informações disponíveis nos mercados financeiros e de capital relevantes, utilizando metodologias e parâmetros consistentes com esses mercados. Modelamos uma série bivariada de número de sinistros (duas linhas de negócios) de seguros financeiros com modelos autorregressivos e utilizamos cópulas para modelar a estrutura de dependência das séries temporais condicionado aos modelos ajustados nas marginais. Com esta abordagem, é possível simular números de sinistros futuros de mais de uma carteira, podendo esse resultado ser utilizado em uma avaliação consistente de provisões e da saúde financeira da seguradora. / [en] The objective of this thesis is to present new alternatives for modeling random variables in the insurance industry, using the framework of the score driven models with time-varying parameters. In the first paper, we propose a dynamic model for the aggregate claims distribution, which corresponds to a random sum of claims severity in a certain period of time. Obtaining the aggregate claims distribution is a classic problem in the Risk Theory and fundamental for premium estimation, measurement of obligations and ruin probability valuation. However, obtaining the analytic expression for this probability distribution is a hard task. In this work, we specify nonGaussian distributions for both the number of claims and for the claims severity, under the GAS framework, and, through the use of the fast Fourier transform, we obtain, numerically, the aggregate claims distribution. The second paper deals with the incorporation of macroeconomic variables on the modeling of relevant variables in the insurance sector, in line with the international requirements for market consistent valuation of insurance liabilities, which means that one should take into account the available information in relevant financial and capital markets, using methodologies and parameters consistent with these markets. We model a bivariate time series (two lines of business) of financial insurance with autoregressive models and use copulas models to consider the dependency structure of the time series conditioned to the fitted models for the marginals. Within this approach, it is possible to simulate the numbers of claims from more than one portfolio, and this result can be used in a consistent valuation of liabilities and of the financial health of an insurer.

Le contenu de cette thèse est divisé de la façon suivante. Après un premier chapitre d’introduction, le Chapitre 2 est consacré à introduire aussi simplement que possible certaines des théories qui seront utilisées dans les deux premiers articles. Dans un premier temps, nous discuterons des points importants pour la construction de l’intégrale stochastique par rapport aux semimartingales avec paramètre spatial. Ensuite, nous décrirons les principaux résultats de la théorie de l’évaluation en monde neutre au risque et, finalement, nous donnerons une brève description d’une méthode d’optimisation connue sous le nom de dualité. Les Chapitres 3 et 4 traitent de la modélisation de l’illiquidité et font l’objet de deux articles. Le premier propose un modèle en temps continu pour la structure et le comportement du carnet d’ordres limites. Le comportement du portefeuille d’un investisseur utilisant des ordres de marché est déduit et des conditions permettant d’éliminer les possibilités d’arbitrages sont données. Grâce à la formule d’Itô généralisée il est aussi possible d’écrire la valeur du portefeuille comme une équation différentielle stochastique. Un exemple complet de modèle de marché est présenté de même qu’une méthode de calibrage. Dans le deuxième article, écrit en collaboration avec Bruno Rémillard, nous proposons un modèle similaire mais cette fois-ci en temps discret. La question de tarification des produits dérivés est étudiée et des solutions pour le prix des options européennes de vente et d’achat sont données sous forme explicite. Des conditions spécifiques à ce modèle qui permettent d’éliminer l’arbitrage sont aussi données. Grâce à la méthode duale, nous montrons qu’il est aussi possible d’écrire le prix des options européennes comme un problème d’optimisation d’une espérance sur en ensemble de mesures de probabilité. Le Chapitre 5 contient le troisième article de la thèse et porte sur un sujet différent. Dans cet article, aussi écrit en collaboration avec Bruno Rémillard, nous proposons une méthode de prévision des séries temporelles basée sur les copules multivariées. Afin de mieux comprendre le gain en performance que donne cette méthode, nous étudions à l’aide d’expériences numériques l’effet de la force et la structure de dépendance sur les prévisions. Puisque les copules permettent d’isoler la structure de dépendance et les distributions marginales, nous étudions l’impact de différentes distributions marginales sur la performance des prévisions. Finalement, nous étudions aussi l’effet des erreurs d’estimation sur la performance des prévisions. Dans tous les cas, nous comparons la performance des prévisions en utilisant des prévisions provenant d’une série bivariée et d’une série univariée, ce qui permet d’illustrer l’avantage de cette méthode. Dans un intérêt plus pratique, nous présentons une application complète sur des données financières. / This thesis is structured as follows. After a first chapter of introduction, Chapter 2 exposes as simply as possible different notions that are going to be used in the two first papers. First, we discuss the main steps required to build stochastic integrals for semimartingales with space parameters. Secondly, we describe the main results of risk neutral evaluation theory and, finally, we give a short description of an optimization method known as duality. Chapters 3 and 4 consider the problem of modelling illiquidity, which is covered by two papers. The first one proposes a continuous time model for the structure and the dynamic of the limit order book. The dynamic of a portfolio for an investor using market orders is deduced and conditions to rule out arbitrage are given. With the help of Itô’s generalized formula, it is also possible to write the value of the portfolio as a stochastic differential equation. A complete example of market model along with a calibration method is also given. In the second paper, written in collaboration with Bruno Rémillard, we propose a similar model with discrete time trading. We study the problem of derivatives pricing and give explicit formulas for European option prices. Specific conditions to rule out arbitrage are also provided. Using the dual optimization method, we show that the price of European options can be written as the optimization of an expectation over a set of probability measures. Chapter 5 contained the third paper and studies a different topic. In this paper, also written with Bruno Rémillard, we propose a forecasting method for time series based on multivariate copulas. To provide a better understanding of the proposed method, with the help of numerical experiments, we study the effect of the strength and the structure of the different dependencies on predictions performance. Since copulas allow to isolate the dependence structure and marginal distributions, we study the impact of different marginal distributions on predictions performance. Finally, we also study the effect of estimation errors on the predictions. In all the cases, we compare the performance of predictions by using predictions based on a bivariate series and predictions based on a univariate series, which allows to illustrate the advantage of the proposed method. For practical matters, we provide a complete example of application on financial data.

Les copulas archimédiennes hiérarchiques ont récemment gagné en intérêt puisqu’elles généralisent la famille de copules archimédiennes, car elles introduisent une asymétrie partielle. Des algorithmes d’échantillonnages et des méthodes ont largement été développés pour de telles copules. Néanmoins, concernant l’estimation par maximum de vraisemblance et les tests d’adéquations, il est important d’avoir à disposition la densité de ces variables aléatoires. Ce travail remplie ce manque. Après une courte introduction aux copules et aux copules archimédiennes hiérarchiques, une équation générale sur les dérivées des noeuds et générateurs internes apparaissant dans la densité des copules archimédiennes hiérarchique. sera dérivée. Il en suit une formule tractable pour la densité des copules archimédiennes hiérarchiques. Des exemples incluant les familles archimédiennes usuelles ainsi que leur transformations sont présentés. De plus, une méthode numérique efficiente pour évaluer le logarithme des densités est présentée. / Nested Archimedean copulas recently gained interest since they generalize the well-known class of Archimedean copulas to allow for partial asymmetry. Sampling algorithms and strategies have been well investigated for nested Archimedean copulas. However, for likelihood based inference such as estimation or goodness-of-fit testing it is important to have the density. The present work fills this gap. After a short introduction on copula and nested Archimedean copulas, a general formula for the derivatives of the nodes and inner generators appearing in nested Archimedean copulas is developed. This leads to a tractable formula for the density of nested Archimedean copulas. Various examples including famous Archimedean families and transformations of such are given. Furthermore, a numerically efficient way to evaluate the log-density is presented.

Cette thèse a pour but le développement de certains aspects de la modélisation de la dépendance dans la gestion des risques en dimension plus grande que un. Le premier chapitre est constitué d'une introduction générale. Le deuxième chapitre est constitué d'un article s'intitulant « Estimating Bivariate Tail : a copula based approach », soumis pour publication. Il concerne la construction d'un estimateur de la queue d'une distribution bivariée. La construction de cet estimateur se fonde sur une méthode de dépassement de seuil (Peaks Over Threshold method) et donc sur une version bivariée du Théorème de Pickands-Balkema-de Haan. La modélisation de la dépendance est obtenue via la Upper Tail Dependence Copula. Nous démontrons des propriétés de convergence pour l'estimateur ainsi construit. Le troisième chapitre repose sur un article: « A multivariate extension of Value-at-Risk and Conditional-Tail-Expectation», soumis pour publication. Nous abordons le problème de l'extension de mesures de risque classiques, comme la Value-at-Risk et la Conditional-Tail-Expectation, dans un cadre multidimensionnel en utilisant la fonction de Kendall multivariée. Enfin, dans le quatrième chapitre de la thèse, nous proposons un estimateur des courbes de niveau d'une fonction de répartition bivariée avec une méthode plug-in. Nous démontrons des propriétés de convergence pour les estimateurs ainsi construits. Ce chapitre de la thèse est lui aussi constitué d'un article, s'intitulant « Plug-in estimation of level sets in a non-compact setting with applications in multivariate risk theory», accepté pour publication dans la revue ESAIM:Probability and Statistics. / In this PhD thesis we consider different aspects of dependence modeling with applications in multivariate risk theory. The first chapter is constituted by a general introduction. The second chapter is essentially constituted by the article “Estimating Bivariate Tail: a copula based approach”, actually submitted for publication. It deals with the problem of estimating the tail of a bivariate distribution function. We develop a general extension of the POT (Peaks-Over-Threshold) method, mainly based on a two-dimensional version of the Pickands-Balkema-de Haan Theorem. The dependence structure between the marginals in the upper tails is described by the Upper Tail Dependence Copula. Then we construct a two-dimensional tail estimator and study its asymptotic properties. The third chapter of this thesis is based on the article “A multivariate extension of Value-at-Risk and Conditional-Tail-Expectation” and submitted for publication. We propose a multivariate generalization of risk measures as Value-at-Risk and Conditional-Tail-Expectation and we analyze the behavior of these measures in terms of classical properties of risk measures. We study the behavior of these measures with respect to different risk scenarios and stochastic ordering of marginals risks. Finally in the fourth chapter we introduce a consistent procedure to estimate level sets of an unknown bivariate distribution function, using a plug-in approach in a non-compact setting. Also this chapter is constituted by the article “Plug-in estimation of level sets in a non-compact setting with applications in multivariate risk theory”, accepted for publication in ESAIM: Probability and Statistics journal.

Este estudo contempla dois ensaios em finanças quantitativas, relacionados, respectivamente, a modelos de apreçamento e risco sistêmico. No Capitulo 1, e apresentado uma alternativa para modelar opções multidimensionais, cujas estruturas de ganhos e perdas dependam das trajetórias dos processos dos preços dos ativos objetos. A modelagem sugerida considera os processos de Levy, uma classe de processos estocásticos bastante ampla, que permite a existência de saltos (descontinuidades) no processo dos preços dos ativos financeiros, e tem como caso particular o movimento Browniano. Para escrever a dependência entre os processos, os conceitos estáticos de copulas ordinárias são estendidos para o contexto dos processos de Levy, levando em consideração a medida de Levy, que caracteriza o comportamento dos saltos. São realizados estudos comparativos entre as copulas dinâmicas de Clayton e de Frank, no apreçamento dos contratos derivativos do tipo asiático, utilizando-se processos gama e técnicas de simulação de Monte Carlo. No Capitulo 2, a estrutura e dinâmica interbancária das exposições mutuas entre as instituições financeiras no Brasil e explorada bem como o capital destas reservas, utilizando um conjunto de dados únicos que considera vários períodos entre 2007 e 2008. Para isto e mostrado que a rede de exposições pode ser modelada adequadamente como um gráfico estocástico dirigido de escala - livre (ponderada) seguindo distribuições que apresentam caudas grossas. A relação entre as conexões das instituições financeiras e seu colchão-de-capital também são investigados neste estudo. Finalmente, a estrutura da rede e usada para explorar a extensão de risco sistêmico gerada no sistema individualmente pelas instituições financeiras. / This study comprises two essays in quantitative finance, related, respectively, to models in asset pricing and systemic risk. In Chapter 1, it is presented an alternative to modeling multidimensional options, where the pay-offs depend on the paths of the trajectories of the underlying-asset prices. The proposed technique considers Levy processes, a very ample class of stochastic processes that allows the existence of jumps (discontinuities) in the price process of financial assets, and as a particular case, comprises the Brownian motion. To describe the dependence among Levy processes, extending the static concepts of the ordinary copulas to the Levy processes context, considering the Levy measure, which characterizes the jumps behavior of these processes. A comparison between the Clayton and the Frank dynamic copulas and their impact in asset pricing of Asian type derivatives contracts is studied, considering gamma processes and Monte Carlo simulation procedures. In Chapter 2, the structure and dynamics of interbank exposures in Brazil using a unique data set of all mutual exposures of financial institutions in Brazil is explored, as well as their capital reserves, at various periods in 2007 and 2008. It is shown that the network of exposures can be adequately modeled as a directed scale-free (weighted) graph with heavy-tailed degree and weight distributions. The relation between connectivity of a financial institution and its capital buffer are also investigated in this study. Finally, the network structure is used to explore the extent of systemic risk generated in the system by the individual institutions.

L'entrée en application depuis le 1er Janvier 2016 de la réforme réglementaire européenne du secteur des assurances Solvabilité 2 est un événement historique qui va changer radicalement les pratiques en matière de gestion des risques. Elle repose sur une prise en compte importante du profil et de la vision du risque, via la possibilité d'utiliser des modèles internes pour calculer les capitaux de solvabilité et l'approche ORSA (Own Risk and Solvency Assessment) pour la gestion interne du risque. La modélisation mathématique est ainsi un outil indispensable pour réussir un exercice réglementaire. La théorie du risque doit être en mesure d'accompagner ce développement en proposant des réponses à des problématiques pratiques, liées notamment à la modélisation des dépendances et aux choix des mesures de risques. Dans ce contexte, cette thèse présente une contribution à l'amélioration de la gestion des risques actuariels. En quatre chapitres nous présentons des mesures multivariées de risque et leurs applications à l'allocation du capital de solvabilité. La première partie de cette thèse est consacrée à l'introduction et l'étude d'une nouvelle famille de mesures multivariées élicitables de risque qu'on appellera des expectiles multivariés. Son premier chapitre présente ces mesures et explique les différentes approches utilisées pour les construire. Les expectiles multivariés vérifient un ensemble de propriétés de cohérence que nous abordons aussi dans ce chapitre avant de proposer un outil d'approximation stochastique de ces mesures de risque. Les performances de cette méthode étant insuffisantes au voisinage des niveaux asymptotiques des seuils des expectiles, l'analyse théorique du comportement asymptotique est nécessaire, et fera le sujet du deuxième chapitre de cette partie. L'analyse asymptotique est effectuée dans un environnement à variations régulières multivariées, elle permet d'obtenir des résultats dans le cas des queues marginales équivalentes. Nous présentons aussi dans le deuxième chapitre le comportement asymptotique des expectiles multivariés sous les hypothèses précédentes en présence d'une dépendance parfaite, ou d'une indépendance asymptotique, et nous proposons à l'aide des statistiques des valeurs extrêmes des estimateurs de l'expectile asymptotique dans ces cas. La deuxième partie de la thèse est focalisée sur la problématique de l'allocation du capital de solvabilité en assurance. Elle est composée de deux chapitres sous forme d'articles publiés. Le premier présente une axiomatisation de la cohérence d'une méthode d'allocation du capital dans le cadre le plus général possible, puis étudie les propriétés de cohérence d'une approche d'allocation basée sur la minimisation d'indicateurs multivariés de risque. Le deuxième article est une analyse probabiliste du comportement de cette dernière approche d'allocation en fonction de la nature des distributions marginales des risques et de la structure de la dépendance. Le comportement asymptotique de l'allocation est aussi étudié et l'impact de la dépendance est illustré par différents modèles marginaux et différentes copules. La présence de la dépendance entre les différents risques supportés par les compagnies d'assurance fait de l'approche multivariée une réponse plus appropriée aux différentes problématiques de la gestion des risques. Cette thèse est fondée sur une vision multidimensionnelle du risque et propose des mesures de nature multivariée qui peuvent être appliquées pour différentes problématiques actuarielles de cette nature / The entry into force since January 1st, 2016 of Solvency 2, the European regulatory reform of insurance industry, is a historic event that will radically change the practices in risk management. It is based on taking into account the own risk profile and the internal view of risk through the ability to use internal models for calculating solvency capital requirement and ORSA (Own Risk and Solvency Assessment) approach for internal risk management. It makes the mathematical modeling an essential tool for a successful regulatory exercise. The risk theory must allow to support this development by providing answers to practical problems, especially those related to the dependence modeling and the choice of risk measures. In the same context, this thesis presents a contribution to improving the management of insurance risks. In four chapters we present multivariate risk measures and their application to the allocation of solvency capital. The first part of this thesis is devoted to the introduction and study of a new family of multivariate elicitable risk measures that we will call multivariate expectiles. The first chapter presents these measures and explains the different construction approaches. The multivariate expectiles verify a set of coherence properties that we also discuss in this chapter before proposing a stochastic approximation tool of these risk measures. The performance of this method is insufficient in the asymptotic levels of the expectiles thresholds. That makes the theoretical analysis of the asymptotic behavior necessary. The asymptotic behavior of multivariate expectiles is then the subject of the second chapter of this part. It is studied in a multivariate regular variations framework, and some results are given in the case of equivalent marginal tails. We also study in the second chapter of the first part the asymptotic behavior of multivariate expectiles under previous assumptions in the presence of a perfect dependence, or in the case of asymptotic independence. Finally, we propose using extreme values statistics some estimators of the asymptotic expectile in these cases. The second part of the thesis is focused on the issue of solvency capital allocation in insurance. It is divided into two chapters; each chapter consists of a published paper. The first one presents an axiomatic characterization of the coherence of a capital allocation method in a general framework. Then it studies the coherence properties of an allocation approach based on the minimization of some multivariate risk indicators. The second paper is a probabilistic analysis of the behavior of this capital allocation method based on the nature of the marginal distributions of risks and the dependence structure. The asymptotic behavior of the optimal allocation is also studied and the impact of dependence is illustrated using some selected models and copulas. Faced to the significant presence of dependence between the various risks taken by insurance companies, a multivariate approach seems more appropriate to build responses to the various issues of risk management. This thesis is based on a multidimensional vision of risk and proposes some multivariate risk measures that can be applied to several actuarial issues of a multivariate nature

A atividade agrícola é uma das atividades mais importantes para o ser humano, pois é fonte de matéria prima, alimentação e energia. No entanto, os principais grandes riscos desta atividade são os de produção e de mercado. A maneira mais popular de gerenciamento destes riscos são os seguros agrícolas. O seguro de produtividade agrícola gerencia o risco de quebra de produção relacionado às adversidades climáticas. Por sua vez, o seguro de faturamento agrícola é uma maneira de gerenciar os riscos de produção e de mercado (preços) conjuntamente. Portanto, este trabalho tem por objetivo apresentar métodos alternativos para precificação dos riscos de produtividade e faturamento agrícola. Na modelagem da distribuição de produtividade utilizou-se distribuições paramétricas que capturam a simetria, a assimetria e a bimodalidade dos dados, características estas geralmente encontradas em produtividades brasileiras. Além disso, a metodologia de cópulas foi utilizada na análise multivariada entre produtividade e preço. Esta metodologia apresenta um ganho significante para estudos de problemas multivariados comparados à distribuição normal multivariada, pois pode-se levar em consideração estruturas de dependências linear, não linear e dependência somente nas caudas da distribuição multivariada. Além disso especifica qualquer tipo de distribuição para as distribuições marginais. Ademais, analisou-se a estrutura de dependência entre produtividade, preços futuros de contrato de soja negociados na Chicago Mercantil Exchange (CME) e a cotação do dólar. Esta análise tridimensional do problema do seguro de faturamento é inédita e foi utilizada pelo fato da soja produzida no Brasil ser exportada e os produtores utilizarem como referência internacional de preços a bolsa americana. Além disso, esta análise tridimensional via cópulas considera a variabilidade e a influência do dólar e o preço futuro da commoditie na modelagem do seguro. Todas as análises foram discutidas e comparadas com as taxas dos seguros aplicadas pelas seguradoras brasileiras. No caso do seguro de produtividade e faturamento (tridimensional), os resultados sugerem que as taxas cobradas pelas seguradoras estão superfaturadas quando comparadas com a metodologia apresentada. A superestimação da taxa dificulta a expressiva venda de seguros no Brasil, além de atrair agricultores com maiores riscos, fortalecendo o problema de seleção adversa. No caso do seguro de faturamento (bidimensional), em que não se leva em consideração a influência do câmbio (dólar) na modelagem, os resultados sugerem que as seguradoras subestimam as taxas do seguro de faturamento, o que pode levar a uma perda grande para a seguradora, pois esta pode estar considerando um risco muito menor do que deveria ser levado em consideração. Na realização das análises foi utilizado o software R. / The agricultural activity is one of the most important activities for humans, as it is a source of raw materials, food and energy. However, this activity presents two major risks: production and market. Agricultural insurance is the most efficient way of managing these risks. Insuring agricultural productivity allows to manage the risk of crop failure due to weather adversities. However, revenue insurance is a way to manage both the production and the market (prices) risks simultaneously. This study aims to present alternative pricing methods risks of agricultural productivity and billing. In modeling the distribution of yield, parametric distributions that capture symmetry, asymmetry and bimodality of the data were used, features usually found in yield in Brazil. In addition, the coupling methodology was used in multivariate analysis between yield and prices. This methodology offers an exceptional gain of multivariate problems compared to the Multivariate Normal distribution, as it takes into account structures of linear and non-linear dependencies and dependency only on the tails of multivariate distribution, in addition to specifying any distribution type for marginal distributions. In addition, we analyzed the dependence structure between yield soybean, future prices traded at the Chicago Mercantil Exchange(CME) and, exchange rate (USD/R$). The three-dimensional analysis of insurance billing via copulas considers variability and influence of dollar and price of commodities exported and traded at CME. All analyses were discussed and compared with insurance rates applied by insurance companies. All analysis are conducted using the R software.

A atividade agrícola é uma das atividades mais importantes para o ser humano, pois é fonte de matéria prima, alimentação e energia. No entanto, os principais grandes riscos desta atividade são os de produção e de mercado. A maneira mais popular de gerenciamento destes riscos são os seguros agrícolas. O seguro de produtividade agrícola gerencia o risco de quebra de produção relacionado às adversidades climáticas. Por sua vez, o seguro de faturamento agrícola é uma maneira de gerenciar os riscos de produção e de mercado (preços) conjuntamente. Portanto, este trabalho tem por objetivo apresentar métodos alternativos para precificação dos riscos de produtividade e faturamento agrícola. Na modelagem da distribuição de produtividade utilizou-se distribuições paramétricas que capturam a simetria, a assimetria e a bimodalidade dos dados, características estas geralmente encontradas em produtividades brasileiras. Além disso, a metodologia de cópulas foi utilizada na análise multivariada entre produtividade e preço. Esta metodologia apresenta um ganho significante para estudos de problemas multivariados comparados à distribuição normal multivariada, pois pode-se levar em consideração estruturas de dependências linear, não linear e dependência somente nas caudas da distribuição multivariada. Além disso especifica qualquer tipo de distribuição para as distribuições marginais. Ademais, analisou-se a estrutura de dependência entre produtividade, preços futuros de contrato de soja negociados na Chicago Mercantil Exchange (CME) e a cotação do dólar. Esta análise tridimensional do problema do seguro de faturamento é inédita e foi utilizada pelo fato da soja produzida no Brasil ser exportada e os produtores utilizarem como referência internacional de preços a bolsa americana. Além disso, esta análise tridimensional via cópulas considera a variabilidade e a influência do dólar e o preço futuro da commoditie na modelagem do seguro. Todas as análises foram discutidas e comparadas com as taxas dos seguros aplicadas pelas seguradoras brasileiras. No caso do seguro de produtividade e faturamento (tridimensional), os resultados sugerem que as taxas cobradas pelas seguradoras estão superfaturadas quando comparadas com a metodologia apresentada. A superestimação da taxa dificulta a expressiva venda de seguros no Brasil, além de atrair agricultores com maiores riscos, fortalecendo o problema de seleção adversa. No caso do seguro de faturamento (bidimensional), em que não se leva em consideração a influência do câmbio (dólar) na modelagem, os resultados sugerem que as seguradoras subestimam as taxas do seguro de faturamento, o que pode levar a uma perda grande para a seguradora, pois esta pode estar considerando um risco muito menor do que deveria ser levado em consideração. Na realização das análises foi utilizado o software R. / The agricultural activity is one of the most important activities for humans, as it is a source of raw materials, food and energy. However, this activity presents two major risks: production and market. Agricultural insurance is the most efficient way of managing these risks. Insuring agricultural productivity allows to manage the risk of crop failure due to weather adversities. However, revenue insurance is a way to manage both the production and the market (prices) risks simultaneously. This study aims to present alternative pricing methods risks of agricultural productivity and billing. In modeling the distribution of yield, parametric distributions that capture symmetry, asymmetry and bimodality of the data were used, features usually found in yield in Brazil. In addition, the coupling methodology was used in multivariate analysis between yield and prices. This methodology offers an exceptional gain of multivariate problems compared to the Multivariate Normal distribution, as it takes into account structures of linear and non-linear dependencies and dependency only on the tails of multivariate distribution, in addition to specifying any distribution type for marginal distributions. In addition, we analyzed the dependence structure between yield soybean, future prices traded at the Chicago Mercantil Exchange(CME) and, exchange rate (USD/R$). The three-dimensional analysis of insurance billing via copulas considers variability and influence of dollar and price of commodities exported and traded at CME. All analyses were discussed and compared with insurance rates applied by insurance companies. All analysis are conducted using the R software.

Les copulas archimédiennes hiérarchiques ont récemment gagné en intérêt puisqu’elles généralisent la famille de copules archimédiennes, car elles introduisent une asymétrie partielle. Des algorithmes d’échantillonnages et des méthodes ont largement été développés pour de telles copules. Néanmoins, concernant l’estimation par maximum de vraisemblance et les tests d’adéquations, il est important d’avoir à disposition la densité de ces variables aléatoires. Ce travail remplie ce manque. Après une courte introduction aux copules et aux copules archimédiennes hiérarchiques, une équation générale sur les dérivées des noeuds et générateurs internes apparaissant dans la densité des copules archimédiennes hiérarchique. sera dérivée. Il en suit une formule tractable pour la densité des copules archimédiennes hiérarchiques. Des exemples incluant les familles archimédiennes usuelles ainsi que leur transformations sont présentés. De plus, une méthode numérique efficiente pour évaluer le logarithme des densités est présentée. / Nested Archimedean copulas recently gained interest since they generalize the well-known class of Archimedean copulas to allow for partial asymmetry. Sampling algorithms and strategies have been well investigated for nested Archimedean copulas. However, for likelihood based inference such as estimation or goodness-of-fit testing it is important to have the density. The present work fills this gap. After a short introduction on copula and nested Archimedean copulas, a general formula for the derivatives of the nodes and inner generators appearing in nested Archimedean copulas is developed. This leads to a tractable formula for the density of nested Archimedean copulas. Various examples including famous Archimedean families and transformations of such are given. Furthermore, a numerically efficient way to evaluate the log-density is presented.