The dissertation consists of two parts. The first part (Chapter 1 to 4) is on some contributions to the development of a non-linear analysis on the quintessential fractal set Sierpinski gasket and its probabilistic interpretation. The second part (Chapter 5) is on the asymptotic tail decays for suprema of stochastic processes satisfying certain conditional increment controls. Chapters 1, 2 and 3 are devoted to the establishment of a theory of backward problems for non-linear stochastic differential equations on the gasket, and to derive a probabilistic representation to some parabolic type partial differential equations on the gasket. In Chapter 2, using the theory of Markov processes, we derive the existence and uniqueness of solutions to backward stochastic differential equations driven by Brownian motion on the Sierpinski gasket, for which the major technical difficulty is the exponential integrability of quadratic processes of martingale additive functionals. A Feynman-Kac type representation is obtained as an application. In Chapter 3, we study the stochastic optimal control problems for which the system uncertainties come from Brownian motion on the gasket, and derive a stochastic maximum principle. It turns out that the necessary condition for optimal control problems on the gasket consists of two equations, in contrast to the classical result on ℝ^d, where the necessary condition is given by a single equation. The materials in Chapter 2 are based on a joint work with Zhongmin Qian (referenced in Chapter 2). Chapter 4 is devoted to the analytic study of some parabolic PDEs on the gasket. Using a new type of Sobolev inequality which involves singular measures developed in Section 4.2, we establish the existence and uniqueness of solutions to these PDEs, and derive the space-time regularity for solutions. As an interesting application of the results in Chapter 4 and the probabilistic representation developed in Chapter 2, we further study Burgers equations on the gasket, to which the space-time regularity for solutions is deduced. The materials in Chapter 4 are based on a joint work with Zhongmin Qian (referenced in Chapter 4). In Chapter 5, we consider a class of continuous stochastic processes which satisfy the conditional increment control condition. Typical examples include continuous martingales, fractional Brownian motions, and diffusions governed by SDEs. For such processes, we establish a Doob type maximal inequality. Under additional assumptions on the tail decays of their marginal distributions, we derive an estimate for the tail decay of the suprema (Theorem 5.3.2), which states that the suprema decays in a manner similar to the margins of the processes. In Section 5.4, as an application of Theorem 5.3.2, we derive the existence of strong solutions to a class of SDEs. The materials in this chapter is based on the work [44] by the author (Section 5.2 and Section 5.3) and an ongoing joint project with Guangyu Xi (Section 5.4).

Nesta tese, apresentamos uma metodologia concreta para calcular os controles -ótimos para sistemas estocásticos não-Markovianos. A análise trajetória a trajetória e o uso da estrutura de discretização proposta por Leão e Ohashi [36] conjuntamente com argumentos de seleção mensuráveis, nos forneceu uma estrutura para transformar um problema infinito dimensional para um finito dimensional. Desta forma, garantimos uma descrição concreta para uma classe bastante geral de problemas. / In this thesis, we present a concrete methodology to calculate the -optimal controls for non-Markovian stochastic systems. A pathwise analysis and the use of the discretization structure proposed by Leão and Ohashi [36] jointly with measurable selection arguments, allows us a structure to transform an infinite dimensional problem into a finite dimensional. In this way, we guarantee a concrete description for a rather general class of stochastic problems.

Processos estocásticos complexos são muitas vezes utilizados em modelagem, com o intuito de capturar uma maior proporção das principais características dos sistemas biológicos. A descrição do comportamento desses sistemas tem sido realizada por muitos amostradores baseados na distribuição a posteriori de Monte Carlo. Modelos probabilísticos que descrevem esses processos podem levar a funções de verossimilhança computacionalmente intratáveis, impossibilitando a utilização de métodos de inferência estatística clássicos e os baseados em amostragem por meio de MCMC. A Computação Bayesiana Aproximada (ABC) é considerada um novo método de inferência com base em estatísticas de resumo, ou seja, valores calculados a partir do conjunto de dados (média, moda, variância, etc.). Essa metodologia combina muitas das vantagens da eficiência computacional de processos baseados em estatísticas de resumo com inferência estatística bayesiana uma vez que, funciona bem para pequenas amostras e possibilita incorporar informações passadas em um parâmetro e formar uma priori para análise futura. Nesse trabalho foi realizada uma comparação entre os métodos de estimação, clássico, bayesiano e ABC, para estudos de simulação de modelos simples e para análise de dados de dinâmica populacional. Foram implementadas no software R as distâncias modular e do máximo como alternativas de função distância a serem utilizadas no ABC, além do algoritmo ABC de rejeição para equações diferenciais estocásticas. Foi proposto sua utilização para a resolução de problemas envolvendo modelos de interação populacional. Os estudos de simulação mostraram melhores resultados quando utilizadas as distâncias euclidianas e do máximo juntamente com distribuições a priori informativas. Para os sistemas dinâmicos, a estimação por meio do ABC apresentou resultados mais próximos dos verdadeiros bem como menores discrepâncias, podendo assim ser utilizado como um método alternativo de estimação. / Complex stochastic processes are often used in modeling in order to capture a greater proportion of the main features of natural systems. The description of the behavior of these systems has been made by many Monte Carlo based samplers of the posterior distribution. Probabilistic models describing these processes can lead to computationally intractable likelihood functions, precluding the use of classical statistical inference methods and those based on sampling by MCMC. The Approxi- mate Bayesian Computation (ABC) is considered a new method for inference based on summary statistics, that is, calculated values from the data set (mean, mode, variance, etc.). This methodology combines many of the advantages of computatio- nal efficiency of processes based on summary statistics with the Bayesian statistical inference since, it works well for small samples and it makes possible to incorporate past information in a parameter and form a prior distribution for future analysis. In this work a comparison between, classical, Bayesian and ABC, estimation methods was made for simulation studies considering simple models and for data analysis of population dynamics. It was implemented in the R software the modular and maxi- mum as alternative distances function to be used in the ABC, besides the rejection ABC algorithm for stochastic differential equations. It was proposed to use it to solve problems involving models of population interaction. The simulation studies showed better results when using the Euclidean and maximum distances together with informative prior distributions. For the dynamic systems, the ABC estimation presented results closer to the real ones as well as smaller discrepancies and could thus be used as an alternative estimation method.

Orientador: Pedro Jose Catuogno / Tese (doutorado) - Universidade Estadual de Campinas, Instituto de Matematica, Estatistica e Computação Cientifica / Made available in DSpace on 2018-08-14T19:44:56Z (GMT). No. of bitstreams: 1 Vieira_MarceloGoncalvesOliveira_D.pdf: 804383 bytes, checksum: ab79ef394c82b721e298a47eaa86c2f6 (MD5) Previous issue date: 2009 / Resumo: Este trabalho aborda homotopias não usuais entre soluções de equações pertencentes a uma coleção de equações. Cada coleção de equações é denominada pelo termo sistema e neste trabalho são considerados dois tipos de sistemas, os sistemas de Young e os sistemas rugosos. Sob determinadas condições, mostramos que um conjunto de pontos acessíveis de um sistema de Young admite recobrimento e um resultado análogo para sistemas rugosos também é válido. Além disso, mostramos que a concatenação de trajetórias de um sistema ainda é uma trajetória deste sistema. Com esse resultado é possível definir uma operação entre as classes de homotopias de trajetórias de um sistema. Outro ponto abordado é estender ao contexto de um sistema de Young a noção de trajetórias regulares de equações diferenciais ordinárias pertencentes a um sistema de controle. Nesta direção obtivemos um resultado o qual diz que a concatenação entre uma trajetória regular e qualquer outra trajetória produz uma trajetória regular. Por fim, estudamos como o conceito de homotopia entre trajetórias de um sistema rugoso se relaciona com conjugação de sistemas e com equações diferenciais estocásticas. / Abstract: This work accosts unusual homotopy between solutions of equations belonging to a collection of equations. Each collection of equations is called by system and in this work are considered two types of systems, Young systems and rough systems. Under certain conditions, we show that a set of points accessible from an Young system admits covering and a similar result for rough systems is also valid. Furthermore, we show that the concatenation of trajectories of a system is also a trajectory of the system. With this result it is possible to define an operation between the classes of homotopy between trajectories of a system. Another point discussed is to extend to the context of trajectories of an Young system the notion of regularity of trajectories of ordinary differential equations belonging to a control system. In this way we obtain a result which says that the concatenation of a regular trajectory and any other trajectory produces a regular trajectory. Finally, we study how the concept of homotopy between trajectories of a rough system relates with conjugation of systems and stochastic differential equations. / Doutorado / Matematica / Doutor em Matemática

M.Sc. (Mathematical Statistics) / Stochastic Differential Equations (SDE’s) are commonly found in most of the modern finance used today. In this dissertation we use SDE’s to model a random phenomenon known as the short-term interest rate where the explanatory power of a particular short-term interest rate model is largely dependent on the description of the SDE to the real data. The challenge we face is that in most cases the transition density functions of these models are unknown and therefore, we need to find reliable and accurate alternative estimation techniques. In this dissertation, we discuss estimating techniques for discretely sampled continuous diffusion processes that do not require the true transition density function to be known. Moreover, the reader is introduced to the following techniques: (i) continuous time maximum likelihood estimation; (ii) discrete time maximum likelihood estimation; and (iii) estimating functions. We show through a Monte Carlo simulation study that the parameter estimates obtained from these techniques provide a good approximation to the estimates obtained from the true transition density. We also show that the bias in the mean reversion parameter can be reduced by implementing the jackknife bias reduction technique. Furthermore, the data analysis carried out on South-African interest rate data indicate strongly that single factor models do not explain the variability in the short-term interest rate. This may indicate the possibility of distinct jumps in the South-African interest rate market. Therefore, we leave the reader with the notion of incorporating jumps into a SDE framework.

Stochastic differential equations (SDE’s) are used to describe systems which are influenced by randomness. Here, randomness is modelled as some external source interacting with the system, thus ensuring that the stochastic differential equation provides a more realistic mathematical model of the system under investigation than deterministic differential equations. The behaviour of the physical system can often be described by probability and thus understanding the theory of SDE’s requires the familiarity of advanced probability theory and stochastic processes. SDE’s have found applications in chemistry, physical and engineering sciences, microelectronics and economics. But recently, there has been an increase in the use of SDE’s in other areas like social sciences, computational biology and finance. In modern financial practice, asset prices are modelled by means of stochastic processes. Thus, continuous-time stochastic calculus plays a central role in financial modelling. The theory and application of interest rate modelling is one of the most important areas of modern finance. For example, SDE’s are used to price bonds and to explain the term structure of interest rates. Commonly used models include the Cox-Ingersoll-Ross model; the Hull-White model; and Heath-Jarrow-Morton model. Since there has been an expansion in the range and volume of interest rate related products being traded in the international financial markets in the past decade, it has become important for investment banks, other financial institutions, government and corporate treasury offices to require ever more accurate, objective and scientific forms for the pricing, hedging and general risk management of the resulting positions. Similar to ordinary differential equations, many SDE’s that appear in practical applications cannot be solved explicitly and therefore require the use of numerical methods. For example, to price an American put option, one requires the numerical solution of a free-boundary partial differential equation. There are various approaches to solving SDE’s numerically. Monte Carlo methods could be used whereby the physical system is simulated directly using a sequence of random numbers. Another method involves the discretisation of both the time and space variables. However, the most efficient and widely applicable approach to solving SDE’s involves the discretisation of the time variable only and thus generating approximate values of the sample paths at the discretisation times. This paper highlights some of the various numerical methods that can be used to solve stochastic differential equations. These numerical methods are based on the simulation of sample paths of time discrete approximations. It also highlights how these methods can be derived from the Taylor expansion of the SDE, thus providing opportunities to derive more advanced numerical schemes. / Dissertation (MSc (Mathematics of Finance))--University of Pretoria, 2007. / Mathematics and Applied Mathematics / MSc / unrestricted

Orientador: Paulo Regis Caron Ruffino / Tese (doutorado) - Universidade Estadual de Campinas, Instituto de Matemática Estatística e Computação Científica / Made available in DSpace on 2018-08-27T10:29:03Z (GMT). No. of bitstreams: 1 Morgado_LeandroBatista_D.pdf: 1320837 bytes, checksum: db1015f01556b3de2b1f7ca1c6bf33d3 (MD5) Previous issue date: 2015 / Resumo: Este trabalho aborda alguns aspectos da teoria de equações diferenciais estocásticas em relação a semimartingales com saltos, suas aplicações na decomposição de fluxos estocásticos em variedades, bem como algumas implicações de natureza geométrica. Inicialmente, em uma variedade munida de distribuições complementares, discutimos o problema da decomposição de fluxos estocásticos contínuos, isto é, gerados por EDE em relação ao movimento Browniano. Resultados anteriores garantem a existência de uma decomposição em difeomorfismos que preservam as distribuições até um tempo de parada. Usando a assim denominada equação de Marcus, bem como uma técnica que denominamos equação 'stop and go', vamos construir um fluxo estocástico próximo ao original, com a propriedade adicional que o fluxo construído pode ser decomposto além do tempo de parada inicial. Em seguida, trataremos da decomposição de fluxos estocásticos no caso descontínuo, isto é, para processos gerados por uma EDE em relação a um semimartingale com saltos. Após uma discussão sobre a existência da decomposição, obtemos as EDEs para as componentes respectivas, a partir de uma extensão que propomos da fórmula de Itô-Ventzel-Kunita. Finalmente, propomos um modelo de equações diferenciais estocásticas com retardo incluindo saltos. A ideia é modelar certos fenômenos em que a informação pode chegar ao receptor por diferentes canais: de forma contínua, mas com retardo, e em tempos discretos, de forma instantânea. Vamos abordar aspectos geométricos relacionados ao tema: transporte paralelo em curvas diferenciáveis com saltos, bem como possibilidade de levantamento de uma solução do nosso modelo de equação para o fibrado de bases de uma variedade diferenciável / Abstract: The main subject of this thesis is the theory of stochastic differential equations driven by semimartingales with jumps. We consider applications in the decomposition of stochastic flows in differentiable manifolds, and geometrical aspects about these equations. Initially, in a differentiable manifold endowed with a pair of complementary distributions, we discuss the decomposition of continuous stochastic flows, that is, flows generated by SDEs driven by Brownian motion. Previous results guarantee that, under some assumptions, there exists a decomposition in diffeomorphisms that preserves the distributions up to a stopping time. Using the so called Marcus equation, and a technique that we call 'stop and go' equation, we construct a stochastic flow close to the original one, with the property that the constructed flow can be decomposed further on the stopping time. After, we deal with the decomposition of stochastic flows in the discontinuous case, that is, processes generated by SDEs driven by semimartingales with jumps. We discuss the existence of this decomposition, and obtain the SDEs for the respective components, using an extension of the Itô-Ventzel-Kunita formula. Finally, we propose a model of stochastic differential equations including delay and jumps. The idea is to describe some phenomena such that the information comes to the receptor by different channels: continuously, with some delay, and in discrete times, instantaneously. We deal with geometrical aspects related with this subject: parallel transport in càdlàg curves, and lifting of solutions of these equations to the linear frame bundle of a differentiable manifold / Doutorado / Matematica / Doutor em Matemática

Dans la première partie de cette thèse, nous obtenons l’existence d’une densité et des estimées gaussiennes pour la solution d’une équation différentielle stochastique rétrograde. C’est une application du calcul de Malliavin et plus particulièrement d’une formule d’I. Nourdin et de F. Viens. La deuxième partie de cette thèse est consacrée à la simulation d’une équation aux dérivées partielles stochastique par une méthode probabiliste qui repose sur la représentation de l’équation aux dérivées partielles stochastique en terme d’équation différentielle doublement stochastique rétrograde, introduite par E. Pardoux et S. Peng. On étend dans ce cadre les idées de F. Zhang et E. Gobet et al. sur la simulation d’une équation différentielle stochastique rétrograde. Dans la dernière partie, nous étudions l’erreur faible du schéma d’Euler implicite pour les processus de diffusion et l’équation de la chaleur stochastique. Dans le premier cas, nous étendons les résultats de D. Talay et L. Tubaro. Dans le second cas, nous étendons les travaux de A. Debussche. / No English summary available.

This thesis studies the role of higher moments, that is moments behind mean and variance, in continuous-time, or diffusion, processes, which are commonly used to model so-called high-frequency data. Thereby, the first part is devoted to the derivation of closed-form expression of general (un)conditional (co)moment formulas of the famous CIR process’s solution. A byproduct of this derivation will be a novel way of proving that the process’s transition density is a noncentral chi-square distribution and that its steady-state law is a Gamma distribution. In the second part, we use these moment formulas to derive a near-exact simulation algorithm to the Heston model, in the sense that our algorithm generates pseudo-random numbers that have the same first four moments as the theoretical diffusion process. We will conduct several in-depth Monte Carlo studies to determine which existing simulation algorithm performs best with respect to these higher moments under certain circumstances. We will conduct the same study for the CIR process, which serves as a diffusion for the latent spot variance in the Heston model. The third part discusses several estimation approaches to the Heston model based on high-frequency data, such as MM, GMM, and (pseudo/quasi) ML. For the GMM approach, we will use the moments derived in the first part as moment conditions. We apply the best methodology to actual high-frequency price series of cryptocurrencies and FIAT stocks to provide benchmark parameter estimates. / Die vorliegende Arbeit untersucht die Rolle von höheren Momenten, also Momente, welche über den Erwartungswert und die Varianz hinausgehen, im Kontext von zeitstetigen Zeitreihenmodellen. Solche Diffusionsprozesse werden häufig genutzt, um sogenannte Hochfrequenzdaten zu beschreiben. Teil 1 der Arbeit beschäftigt sich mit der Herleitung von allgemeinen und in geschlossener Form verfügbaren Ausdrücken der (un)bedingten (Ko-)Momente der Lösung zum CIR-Prozess. Mittels dieser Formeln wird auf einem alternativen Weg bewiesen, dass die Übergangsdichte dieses Prozesses mithilfe einer nichtzentralen Chi-Quadrat-Verteilung beschrieben werden kann, und dass seine stationäre Verteilung einer Gamma-Verteilung entspricht. Im zweiten Teil werden die zuvor entwickelten Ausdrücke genutzt, um einen nahezu exakten Simulationsalgorithmus für das Hestonmodell herzuleiten. Dieser Algorithmus ist in dem Sinne nahezu exakt, dass er Pseudo-Zufallszahlen generiert, welche die gleichen ersten vier Momente besitzen, wie der dem Hestonmodell zugrundeliegende Diffusionsprozess. Ferner werden Monte-Carlo-Studien durchgeführt, die ergründen sollen, welche bereits existierenden Simulationsalgorithmen in Hinblick auf die ersten vier Momente die besten Ergebnisse liefern. Die gleiche Studie wird außerdem für die Simulation des CIR-Prozesses durchgeführt, welcher im Hestonmodell als Diffusion für die latente, instantane Varianz dient. Im dritten Teil werden mehrere Schätzverfahren für das Hestonmodell, wie das MM-, GMM und pseudo- beziehungsweise quasi-ML-Verfahren, diskutiert. Diese werden unter Benutzung von Hochfrequenzdaten studiert. Für das GMM-Verfahren dienen die hergeleiteten Momente aus dem ersten Teil der Arbeit als Momentebedingungen. Um ferner Schätzwerte für das Hestonmodell zu finden, werden die besten Verfahren auf Hochfrequenzmarktdaten von Kryptowährungen, sowie hochliquider Aktientitel angewandt. Diese sollen zukünftig als Orientierungswerte dienen.

Monte Carlo path simulations are common in mathematical and computational finance as a way of estimating the expected values of a quantity such as a European put option, which is functional to the solution of a stochastic differential equation (SDE). The computational complexity of the standard Monte Carlo (MC) method grows quite large quickly, so in this thesis we focus on the Multilevel Monte Carlo (MLMC) method by Giles, which uses multigrid ideas to reduce the computational complexity. We use a Euler-Maruyama time discretisation for the approximation of the SDE and investigate how the convergence rate of the MLMC method improves the computational times and cost in comparison with the standard MC method. We perform a numerical analysis on the computational times and costs in order to achieve the desired accuracy and present our findings on the performance of the MLMC method on a European put option compared to the standard MC method.