Commodities futures trading went through unparalleled structural transformation during the first decade of the 2000s, which ultimately resulted in long lasting impacts on the volume and open interest levels as well as on the access to these markets and inclusion of new participants. Benefiting from the new sets of high frequency data made available due to these transformations, this dissertation is composed of three papers that investigate different market microstructure aspects of the commodities futures markets at BM&F-Bovespa. The first paper analyzes the modelling and forecasting of realized volatility in the corn and live cattle markets. For this purpose, the heterogeneous autoregressive model (HAR-RV) proposed by Corsi (2009) was used, as well as its extensions adapted to include jump components (Andersen et al., 2007) and leverage components (Corsi and Reno, 2012). Using measurements to compare both analysis, results show that modelling in-sample realized volatility is best performed if jumps and leverage components are included in the model. Out-of-sample forecasts results for the live cattle market show that there was no statistically significant difference between the models. For the corn markets, difference in the models\' forecast performance was found at the daily horizon only. The second paper delves into the relationship between volatility, volume and bid-ask spread in the corn and live cattle markets. Considering that these are emerging agricultural markets, concentration measures were also included. A three-equation structural model was used to capture the relationship between volatility, volume and bid-ask spread and the estimation was performed using the IV-GMM approach. Findings show that bid-ask spread levels are higher for live cattle markets than it is for corn markets. In addition, bid-ask spread is negatively related to volume and positively related to volatility. The significance and magnitude of the responses depend on the level of liquidity in each market. Further, results point out that concentration impacts corn and live cattle differently. The third paper examines the dynamic relationship between dealers activity and market structure in the live cattle inter-dealer market at BM&F-Bovespa. First, a descriptive analysis of the live cattle inter-dealer market structure is carried out and then follows an investigation of the dynamic of dealers\' activity and its determinants. Results indicate that the live cattle inter-dealer market is not competitive and that dealers\' activity is positively related to market concentration, quoted bid-ask spread, number of active dealers and the dealer\'s traded quantity. / Negociações nos mercados futuros de commodities passaram por transformações estruturais significativas durante a primeira década dos anos 2000, resultando em uma elevação dos níveis de volume e open interest, e também em uma maior facilidade de acesso a esses mercados e inclusão de novos participantes. Beneficiando-se da divulgação de dados de alta frequência possibilitada por estas transformações, esta tese, composta por três artigos, tem por objetivo investigar diferentes aspectos da microestrutura dos mercados de commodities da BM&F-Bovespa. O primeiro artigo analisa a modelagem e previsão de volatilidade realizada nos mercados futuros de milho e boi gordo. Para este fim, utilizou-se o modelo heterogêneo auto regressivo proposto por Corsi (2009), bem como suas extensões adaptadas para a inclusão dos componentes de saltos (jumps) (Andersen et al., 2007) e alavancagem (Corsi e Reno, 2012). Utilizando diferentes métricas de comparação, os resultados encontrados mostram que os modelos que incluem os componentes de saltos e os de alavancagem tem melhor desempenho que os demais em análises in-sample (modelagem). Por outro lado, a análise das previsões out-of-sample mostra que, para o mercado de boi gordo, não há diferença entre os modelos empregados, enquanto que para o mercado de milho verificou-se uma diferenciação preditiva no horizonte diário, porém para os horizontes semanal e mensal, os quatro modelos tiveram performance indistinta. O segundo artigo explora a relação entre volatilidade, volume e bid-ask spread nos mercados de milho e boi gordo. Levando em conta que se trata de mercados emergentes, métricas de concentração de mercado foram incluídas na análise. Para capturar a relação entre volatilidade, volume e bid-ask spread, um modelo estrutural de três equações simultâneas foi utilizado e a estimação foi feita através do modelo GMM com variáveis instrumentais. Os resultados indicam que os níveis de bid-ask spread encontrados para o mercado de boi gordo são maiores que os encontrados para o mercado de milho. Além disso, o bid-ask spread é negativamente relacionado ao volume e positivamente relacionado à volatilidade. Entretanto, a intensidade e magnitude da relação entre as variáveis depende dos níveis de liquidez dos mercados analisados. A concentração impacta o mercado de milho e boi gordo de forma diferente. O terceiro artigo investiga a dinâmica da relação entre a atividade dos dealers e estrutura do mercado de boi gordo na BM&F-Bovespa. Primeiramente, faz-se uma análise descritiva deste mercado e posteriormente estuda-se o comportamento dos dealers e seus determinantes. Os resultados indicam que os dealers no mercado de boi gordo não operam em uma estrutura competitiva e que a atividade destes é positivamente relacionada à concentração de mercado, ao bid-ask spread, ao número de dealers ativos e à quantidade de contratos transacionada pelos dealers.

Atualmente, o circuito de moagem da Votorantim Metais Zinco Morro Agudo (VMZMA) inclui dois moinhos de bolas com capacidade combinada média de 145 t/h, sendo 95 t/h processadas na Linha 1 e 50 t/h na Linha 2. A alimentação de ambas as linhas provém de uma única pilha pulmão, formada pela fração passante em peneira de 10 mm, resultante dos estágios de britagem. As duas linhas de moagem operam de forma independente, cada uma sob circuito fechado com as respectivas baterias de ciclones de 10 polegadas de diâmetro. O produto nominal do circuito deve apresentar 75% passantes na malha 44 m. A operação atual do circuito apresenta carga circulante elevada, da ordem de 400%, bem como eventos regulares de curto-circuito (by-pass) de grossos para o overflow dos ciclones, o que resulta em produto mais grosso do que o estipulado para a moagem. Com o intuito de analisar a eficiência de moagem foram realizadas campanhas de amostragem no circuito industrial da VMZMA, cujo resultado apontou para um melhor desempenho de um dos moinhos. Os dados obtidos por meio das amostragens e de caracterização tecnológica serviram de base para a calibração dos modelos matemáticos dos equipamentos de processo. Simulações indicaram potencial elevado de aumento de capacidade do circuito mediante o emprego de peneiras de alta frequência em substituição ou combinação com ciclones existentes. Recomenda-se avaliar o potencial de melhoria de desempenho metalúrgico no processo de flotação de produtos de peneiramento, ante a produtos de ciclones. / Votorantim Metais Zinco Morro Agudo (VMZMA) grinding circuit has a combined capacity of 145 t/h i.e. 95 t\\h in Line 1 and 50 t/h in Line 2. The feed comprises crushed 10 mm obtained in a three-staged crushing circuit. The two ball mills operate in independent parallel lines configured in a closed circuit with 10 inch cyclones, resulting in a 75% passing in 0.044 mm nominal products. The operation shows a high recirculating load (~400%), as well as frequent oversize by pass to the products, which results in a relatively coarse product. Survey campaigns were conducted in the grinding circuit for assessing its overall performance, as well as for fitting the corresponding mathematical models. The results indicated a better performance associated with one of the two existing ball mill lines. The simulations indicate a high potential for increasing the circuit throughput by introducing high frequency screens in substitution or in combination with the existing cyclones. Further studies were recommended for assessing the metallurgical performance associated with the product, as compared with cyclone products.

Para verificar a contribuição da audiometria em altas freqüências (9 a 20 kHz) e das emissões otoacústicas - produto de distorção para a identificação de alterações auditivas em indivíduos com desordem pigmentar do tipo vitiligo, 30 indivíduos portadores de vitiligo foram avaliados audiologicamente, e comparados a um grupo controle. Tanto para as altas freqüências (9 a 20 kHz) como para as emissões otoacústicas foram observadas diferenças significantes entre os grupos, que não foram evidenciadas na audiometria convencional (0,25 a 8 kHz). Os resultados sugerem prejuízo nas funções auditivas para o grupo com desordem pigmentar do tipo vitiligo. / In order to verify the contribution of the high frequency audiometry (9 to 20 kHz) and distortion product- otoacoustic emissions for the identification of auditory impairments in individuals with pigmentary disorder type vitiligo, 30 individuals with vitiligo were evaluated, and compared with a control group. As much for the high frequencies (9 the 20 kHz) as for the otoacoustic emissions, significant differences were observed between the groups, that were not evidenced in the conventional audiometry (0,25 the 8 kHz). The results suggest damage in the auditory functions for the group with pigmentary disorder type vitiligo

Up to 60% of the World’s oil is now within carbonates, with over 50% in the Middle East. Many existing carbonate fields have very low oil recoveries due to multiple scales of pore heterogeneity. To secure better recoveries the controls from deposition and diagenesis towards the origin of carbonate pore heterogeneity needs better understanding. To provide good sample support, three High frequency Cycle’s were sampled (2 from the Lekhwair Formation and the third being the Lower Kharaib Formation) from an offshore field (Abu Dhabi) along a southwest-northeast transect, encompassing the oil leg, transition zone, water leg, the field crest and two opposing flanks. With respect to deposition, the 4th order Sequence Boundaries’ (hardgrounds) and the Maximum Flooding Surface’s were correlated across the field, within the sequence stratigraphic framework, showing that each HFC, of the Lekhwair Formation, contains laterally continuous reservoirs (4th order HST’s) which are compartmentalised above and below by impermeable seals (4th order TST’s). The Lower Kharaib Formation shows significant shoaling producing the shallowest platform (prolonged 3rd order TST) and the best connected reservoir facies. With respect to diagenesis, δ 18O isotopes trends, from calcite cement zones within macrocements from the water and oil legs, in comparison with oil inclusion abundances suggest that oil charge reduced cementation in the crest macropores. Stylolitisation in the water leg at deep burial provided solutes for new cement nucleation causing near complete macropore occlusion. The most open micropore networks coincide with the highest porosity/permeability relationships at the mid-late HST’s of each HFC. Considering these areas could be lower grade reservoirs, and that pore characterisation by Lucia (1999) does not include identifying and quantifying micropores, a new ‘Micropore model’ (using elements from the Petrotype atlas method) is devised. This new method highlights micropore-dominated areas alongside macropore-dominated areas within specific reservoir horizons. This provides information of pore heterogeneity at several scales within a carbonate reservoir and may determine the method for oil extraction and increase oil recovery from both the Lekhwair and Lower Kharaib Formations.

In this work, we study the backreaction of high-frequency gravitational waves on cosmological backgrounds. To describe the wave, we use the Isa- acson formalism, specifically the WKB approximation, which allows us to express the backreaction through an effective stress-energy tensor of the gra- vitational wave. First, we consider the inhomogeneous cosmological model of Charach and Malin, which contains gravitational waves and a massless scalar field mini- mally coupled to gravity. We show that although this is a spatially compacti- fied solution, it is possible to add a high-frequency perturbation and solve Einstein's equations with the effective stress-energy tensor in a consistent way. The bacreaction is of the same order as the influence of the scalar field. Second, we add multiple incoherent high-frequency waves to the homoge- neous Kasner background, and discuss the relation to the late-time limit of the Gowdy (vacuum Charach and Malin) model. 1

Kyoto University (京都大学) / 0048 / 新制・課程博士 / 博士(工学) / 甲第16855号 / 工博第3576号 / 新制||工||1540(附属図書館) / 29530 / 京都大学大学院工学研究科電気工学専攻 / (主査)教授 引原 隆士, 教授 木本 恒暢, 教授 小野寺 秀俊 / 学位規則第4条第1項該当

Abstract The most important paradigms in quantum mechanics are probably a twolevel system, a harmonic oscillator and an ideal (infinite) periodic potential. The first two provide a starting point for understanding the phenomena in systems where the spectrum of energy levels is discrete, whereas the last one results in continuous energy bands. Here an attempt is made to study the dynamics of the electrons in a narrow miniband of a semiconductor superlattice under electric and magnetic fields. Semiconductor superlattices are artificial periodic structures, where certain properties like the period and the energy band structure, defined in standard crystals by the nature, can be controlled. Electron dynamics in a single superlattice miniband is interesting both from the viewpoint of fundamental and applied physics. From the fundamental perspective superlattices serve as a model system for a wealth of phenomena resulting from the wavenature of charge carriers. On the other hand, superlattices can potentially be utilized in oscillators and amplifiers operating at THz frequencies. They can, in principle, provide a reasonable THz Bloch gain under dc bias and parametric amplification in the presence of ac pump field. Because of numerous scientific and technological applications in different areas of science and technology, including astrophysics and atmospheric science, biological and medical sciences, and detection of concealed weapons and biosecurity, a construction of compact tunable THz amplifiers and generators that can operate at room temperature is an important – but so far unrealized – task. This thesis focuses on the influence of electric and magnetic fields on small-signal absorption and gain in semiconductor superlattices in the presence of dissipation (scattering). We present several new ideas how the effects arising due to the wave nature of the electrons can be utilized in an operation of THz oscillators and amplifiers. In Papers I–V, we discuss the properties of superlattice sub-THz and THz parametric amplifiers, whereas the Papers VI–IX are devoted to the problem of domain instability in the realization of cw THz Bloch oscillator. In Paper IX we also establish a feasibility of new type of superlattice THz amplifier based on nonlinear cyclotron-like oscillations of the miniband electrons. The ideas presented in the Papers I–IX are supplemented here with a detailed discussion of the physical origin of the effects and more rigorous mathematical derivations of the main equations.

Using high frequency data for forecasting or nowcasting, we have to deal with three major problems: the mixed frequency problem, the high dimensionality (fat re- gression, parameter proliferation) problem, and the unbalanced data problem (miss- ing observations, ragged edge data). We propose a BSTS-U-MIDAS model (Bayesian Structural Time Series-Unlimited-Mixed-Data Sampling model) to handle these prob- lem. This model consists of four parts. First of all, a structural time series with regressors model (STM) is used to capture the dynamics of target variable, and the regressors are chosen to boost the forecast accuracy. Second, a MIDAS model is adopted to handle the mixed frequency of the regressors in the STM. Third, spike- and-slab regression is used to implement variable selection. Fourth, Bayesian model averaging (BMA) is used for nowcasting. We use this model to nowcast quarterly GDP for Canada, and find that this model outperform benchmark models: ARIMA model and Boosting model, in terms of MAE (mean absolute error) and MAPE (mean absolute percentage error). / Graduate / 0501 / 0508 / 0463 / jonduan@uvic.ca

High efficiency of power converters placed between renewable energy sources and the utility grid is required to maximize the utilization of these sources. Power quality is another aspect that requires large passive elements (inductors, capacitors) to be placed between these sources and the grid. The main objective is to develop higher-level high frequency-based power converter system (HFPCS) that optimizes the use of hybrid renewable power injected into the power grid. The HFPCS provides high efficiency, reduced size of passive components, higher levels of power density realization, lower harmonic distortion, higher reliability, and lower cost. The dynamic modeling for each part in this system is developed, simulated and tested. The steady-state performance of the grid-connected hybrid power system with battery storage is analyzed. Various types of simulations were performed and a number of algorithms were developed and tested to verify the effectiveness of the power conversion topologies. A modified hysteresis-control strategy for the rectifier and the battery charging/discharging system was developed and implemented. A voltage oriented control (VOC) scheme was developed to control the energy injected into the grid. The developed HFPCS was compared experimentally with other currently available power converters. The developed HFPCS was employed inside a microgrid system infrastructure, connecting it to the power grid to verify its power transfer capabilities and grid connectivity. Grid connectivity tests verified these power transfer capabilities of the developed converter in addition to its ability of serving the load in a shared manner. In order to investigate the performance of the developed system, an experimental setup for the HF-based hybrid generation system was constructed. We designed a board containing a digital signal processor chip on which the developed control system was embedded. The board was fabricated and experimentally tested. The system’s high precision requirements were verified. Each component of the system was built and tested separately, and then the whole system was connected and tested. The simulation and experimental results confirm the effectiveness of the developed converter system for grid-connected hybrid renewable energy systems as well as for hybrid electric vehicles and other industrial applications.

In this dissertation, I address a range of topics in the context of mutual fund performance and high frequency trading. The first chapter provides novel evidence on the role of ambiguity aversion in determining the response of mutual fund investors to historical fund performance information. It presents a model of ambiguity averse investors who receive multiple performance-based signals of uncertain precision about manager skill. A key implication of the model is that when investors receive multiple signals of uncertain quality, they place a greater weight on the worst signal. There is strong empirical support for this prediction in the data. Fund flows display significantly higher sensitivity to the worst performance measure even after controlling for fund performance at multiple horizons, performance volatility, flow-performance convexity, and a host of other relevant explanatory variables. This effect is particularly pronounced in the case of retail funds in contrast to institutional funds. The results suggest that fund investor behavior is best characterized as reflecting both Bayesian learning and ambiguity aversion. The second chapter combines data on high frequency trading (HFT) activities of a randomly selected sample of 120 stocks and data on institutional trades, I find that HFT increases the trading costs of traditional institutional investors. An increase of one standard deviation in the intensity of HFT activities increases institutional execution shortfall costs by a third. Further analysis suggests that HFT represents an opportunistic and extra-expensive source of liquidity when demand and supply among institutional investors are imbalanced. Moreover, the impact on institutional trading costs is most pronounced when high frequency (HF) traders engage in directional strategies (e.g., momentum ignition and order anticipation). I perform various analyses to rule out an alternative explanation that HF traders are attracted to stocks that have high trading costs. First, HFT is most prevalent in liquid stocks. Second, the results are robust to controls for stable stock liquidity characteristics and events that might jointly affect HFT and trading costs. Third, an analysis of the HFT behavior around the temporary short selling ban in September 2008 highlights the opportunistic nature of liquidity provision by HF traders. Finally, Granger causality tests show that intensive HFT activity significantly contributes to institutional trading costs, but not vice versa. The third chapter analyzes the implications of the tournament-like competition in the mutual fund industry using a framework that addresses the risk-taking incentives facing fund managers. The theoretical model presented in this chapter suggests that the increase in the \emph{activeness} of the interim loser manager's portfolio is directly related to the magnitude of the performance gap at the interim stage, and to the strength of the investor (cash flow) response to the relative performance rankings of the funds (i.e., the strength of the tournament effect). The empirical evidence based on quarterly Active Share data for a sample of domestic stock funds, is consistent with the key predictions of the model.