Liquidity and portfolio optimisation

This thesis presents research within empirical financial economics with focus on liquidity and portfolio optimisation in the stock market. The discussion on liquidity is focused on measurement issues, including TAQ data processing and measurement of systematic liquidity factors (FSO). Furthermore, a framework for treatment of the two topics in combination is provided. The liquidity part of the thesis gives a conceptual background to liquidity and discusses several different approaches to liquidity measurement. It contributes to liquidity measurement by providing detailed guidelines on the data processing needed for applying TAQ data to liquidity research. The main focus, however, is the derivation of systematic liquidity factors. The principal component approach to systematic liquidity measurement is refined by the introduction of moving and expanding estimation windows, allowing for time-varying liquidity co-variances between stocks. Under several liability specifications, this improves the ability to explain stock liquidity and returns, as compared to static window PCA and market average approximations of systematic liquidity. The highest ability to explain stock returns is obtained when using inventory cost as a liquidity measure and a moving window PCA as the systematic liquidity derivation technique. Systematic factors of this setting also have a strong ability in explaining a cross-sectional liquidity variation. Portfolio optimisation in the FSO framework is tested in two empirical studies. These contribute to the assessment of FSO by expanding the applicability to stock indexes and individual stocks, by considering a wide selection of utility function specifications, and by showing explicitly how the full-scale optimum can be identified using either grid search or the heuristic search algorithm of differential evolution. The studies show that relative to mean-variance portfolios, FSO performs well in these settings and that the computational expense can be mitigated dramatically by application of differential evolution.

Projeto iterativo, simulação, análise e otimização de máquina a relutância variável monofásica

Variable reluctance motors have been increasingly used as an alternative for variable speed and high speed drives in many industrial applications, due to many advantages like the simplicity of construction, robustness, and low cost. The most common applications in recent years are related to aeronautics, electric and hybrid vehicles and wind power generation. This paper explores the theory, operation, design procedures and analysis of a variable reluctance machine. An iterative design methodology is introduced and used to design a 1.25 kW prototype. For the analysis of the machine two methods are used, an analytical method and the finite element simulation. The results obtained by both methods are compared. The results of finite element simulation are used to determine the inductance profiles and torque of the prototype. The magnetic saturation is examined visually and numerically in four critical points of the machine. The data collected in the simulation allow the verification of design and operating limits for the prototype. Moreover, the behavior of the output quantities is analyzed (inductance, torque and magnetic saturation) by variation of physical dimensions of the motor. Finally, a multiobjective optimization using Differential Evolution algorithms and Genetic Algorithms for switched reluctance machine design is proposed. The optimized variables are rotor and stator polar arcs, and the goals are to maximize the average torque, the average torque per copper losses and the average torque per core volume. Finally, the initial design and optimized design are compared.

Estratificação de solos em camadas horizontais utilizando evolução diferencial

This work's objective is the development of a methodology to represent an unknown soil through a stratified horizontal multilayer soil model, from which the engineer may carry out eletrical grounding projects with high precision. The methodology uses the experimental electrical apparent resistivity curve, obtained through measurements on the ground, using a 4-wire earth ground resistance tester kit, along with calculations involving the measured resistance. This curve is then compared with the theoretical electrical apparent resistivity curve, obtained through calculations over a horizontally strati ed soil, whose parameters are conjectured. This soil model parameters, such as the number of layers, in addition to the resistivity and the thickness of each layer, are optimized by Differential Evolution method, with enhanced performance through parallel computing, in order to both apparent resistivity curves get close enough, and it is possible to represent the unknown soil through the multilayer horizontal soil model fitted with optimized parameters. In order to assist the Differential Evolution method, in case of a stagnation during an arbitrary amount of generations, an optimization process unstuck methodology is proposed, to expand the search space and test new combinations, allowing the algorithm to nd a better solution and/or leave the local minima. It is further proposed an error improvement methodology, in order to smooth the error peaks between the apparent resistivity curves, by giving opportunities for other more uniform solutions to excel, in order to improve the whole algorithm precision, minimizing the maximum error. Methodologies to verify the polynomial approximation of the soil characteristic function and the theoretical apparent resistivity calculations are also proposed by including middle points among the approximated ones in the verification. Finally, a statistical evaluation prodecure is presented, in order to enable the classication of soil samples. The soil stratification methodology is used in a control group, formed by horizontally stratified soils. By using statistical inference, one may calculate the amount of soils that, within an error margin, does not follow the horizontal multilayer model.

Algoritmos de otimização e o problema inverso aplicados à identificação de danos estruturais com o uso do MEF

Dissertação (mestrado)—Universidade de Brasília, Faculdade de Tecnologia, Departamento de Engenharia Civil e Ambiental, 2015.

Scheduling of head-dependent cascaded hydro systems: Mixed-integer quadratic programming approach

This paper is on the problem of short-term hydro, scheduling, particularly concerning head-dependent cascaded hydro systems. We propose a novel mixed-integer quadratic programming approach, considering not only head-dependency, but also discontinuous operating regions and discharge ramping constraints. Thus, an enhanced short-term hydro scheduling is provided due to the more realistic modeling presented in this paper. Numerical results from two case studies, based on Portuguese cascaded hydro systems, illustrate the proficiency of the proposed approach.

A remote virtual experiment in mechanical engineering

Conferência: 2nd Experiment at International Conference (Exp at)- Univ Coimbra, Coimbra, Portugal - Sep 18-20, 2013

A hybrid procedure to identify the optimal stiffness coefficients of elastically restrained beams

The formulation of a bending vibration problem of an elastically restrained Bernoulli-Euler beam carrying a finite number of concentrated elements along its length is presented. In this study, the authors exploit the application of the differential evolution optimization technique to identify the torsional stiffness properties of the elastic supports of a Bernoulli-Euler beam. This hybrid strategy allows the determination of the natural frequencies and mode shapes of continuous beams, taking into account the effect of attached concentrated masses and rotational inertias, followed by a reconciliation step between the theoretical model results and the experimental ones. The proposed optimal identification of the elastic support parameters is computationally demanding if the exact eigenproblem solving is considered. Hence, the use of a Gaussian process regression as a meta-model is addressed. An experimental application is used in order to assess the accuracy of the estimated parameters throughout the comparison of the experimentally obtained natural frequency, from impact tests, and the correspondent computed eigenfrequency.

Mortality by Heart Failure and Ischemic Heart Disease in Brazil from 1996 to 2011

Background:Circulatory system diseases are the first cause of death in Brazil.Objective:To analyze the evolution of mortality caused by heart failure, by ischemic heart diseases and by ill-defined causes, as well as their possible relations, in Brazil and in the geoeconomic regions of the country (North, Northeast, Center-West, South and Southeast), from 1996 to 2011.Methods:Data were obtained from DATASUS and death declaration records with codes I20 and I24 for acute ischemic diseases, I25 for chronic ischemic diseases, and I50 for heart failure, and codes in chapter XIII for ill-defined causes, according to geoeconomic regions of Brazil, from 1996 to 2011.Results:Mortality rates due to heart failure declined in Brazil and its regions, except for the North and the Northeast. Mortality rates due to acute ischemic heart diseases increased in the North and Northeast regions, especially from 2005 on; they remained stable in the Center-West region; and decreased in the South and in the Southeast. Mortality due to chronic ischemic heart diseases decreased in Brazil and in the Center-West, South and Southeast regions, and had little variation in the North and in the Northeast. The highest mortality rates due to ill-defined causes occurred in the Northeast until 2005.Conclusions:Mortality due to heart failure is decreasing in Brazil and in all of its geoeconomic regions. The temporal evolution of mortality caused by ischemic heart diseases was similar to that of heart failure. The decreasing number of deaths due to ill-defined causes may represent the improvement in the quality of information about mortality in Brazil. The evolution of acute ischemic heart diseases ranged according to regions, being possibly confused with the differential evolution of ill-defined causes.

Estudi comparatiu de l'algorisme basat en una colònia artificial d'abelles

La finalitat d'aquest projecte és la realització d'un estudi comparatiu de l'algoritme basat en una colònia artificial d'abelles, Artificial Bee Colony (ABC), comparat amb un conjunt d'algoritmes fonamentats en el paradigma de la computació evolutiva. S'utilitzarà l'eficàcia a l'hora d'optimitzar diverses funcions com a mesura comparativa. Els algoritmes amb els quals es comparara l'algoritme ABC són: algoritmes genètics, evolució diferencial i optimització amb eixam de partícules.

Population differentiation and its consequences in the white campion "Silene latifolia" (Caryophyllaceae): a functional analysis

Summary In his theory On the Origin of Species by Means of Natural Selection (1859), Darwin describes evolution as a gradual change in population over time and that natural selection is a process that caused evolution. Because quantitative variation in species is partly influenced by several genes and thus heritable, association between levels of genetic variation at neutral markers and at quantitative traits and their partitioning within and among populations are important to study mechanisms that drive evolution in populations. Most studies addressing quantitative variation in plants focused on morphological and life history traits but not in traits affecting reproductive success. The aim of this thesis is to better understand how patterns of variation for neutral molecular markers and phenotypic traits drive the evolution of reproduction and defensive mechanisms in six European populations of Silene latifolia, a dioecious plant species. We found evidence for extremely high within and between population variation at six microsatellite loci and at most quantitative traits studied in plants grown under standardized conditions (morphology, life history and reproductive traits). Interestingly, there was clinal variation between age at first flowering and latitude. This pattern is likely due to natural selection since differentiation of this trait was high, heritable and probably higher than differentiation at neutral markers. Our study focused on sex specific selective pressures: mechanisms of intersexual coadaptation and defence mechanism against the seed predator Hadena bicruris. To address divergence at reproductive traits, we studied male and female population of origin effects and in particular pollen competitive ability on male post-pollination success in the study populations with within and between populations crosses. We crossed the same female plant with pollen from a male within the same population of origin and pollen from two males from two distinct populations, using a fixed tester male as a competitor. Additionally, we conducted control crosses with pollen from each male as a single donor. We analysed paternity success of each competitor with two microsatellite loci, seed set and offspring fitness. Male population of origin showed significant among-population variation for siring success at pollen competition. In vitro pollen germination rate showed heritable variation among populations and was positively correlated to siring success. Local or foreign pollen did not have a consistent advantage. Furthermore, female population of origin affected the outcome of pollen competition in some populations. There was no difference of seed set or offspring fitness in within/ between population crosses. This suggests that reproductive divergence may occur via pollen competition in Silene latifolia. The specialist seed predator Hadena bicruris may also induce divergence between populations. We tested potential constitutive and induced defence mechanisms against the specialist predator Hadena bicruris. Because fruit wall thickness is smaller in the invasive range (Northern America) were the moth is absent, this suggests that a thicker fruit wall is a potentially defensive trait against larval attack, and that relaxed selection in the absence of the seed predator has resulted in an evolutionary loss of this defence in the invasive range. Fruit wall thickness was different among three populations. Experimental exposure to moth eggs increased fruit abortion. Fruits built after attack on exposed plants did not have thicker fruit walls compared to fruits on non-exposed plants. Furthermore, fruits with thicker fruit walls were not less profitable, nor did they require longer handling time when exposed to larvae, suggesting no defensive role of fruit wall thickness. Our results show that there is high molecular and phenotypic variation in Silene latifolia and that traits potentially involved in reproductive success both for intra-specific (between sexes) and inter-specific interactions are heritable. Different selective forces may thus interact and cause differential evolution of geographically separated Silene latifolia populations in Europe, leading to the observed differentiation. Résumé Dans sa théorie de l'évolution, L'origine des espèces, ch. 4 (1859), Darwin décrit l'évolution comme un processus continu au cours du temps à l'intérieur de populations et que la sélection naturelle en est le moteur. La variation quantitative est en partie déterminée par plusieurs gènes, donc transmissible à la descendance. Associer le niveau de variation génétique à des marqueurs neutres au niveau de la variation à des traits quantitatifs, ainsi que la répartition à l'intérieur et entre les populations d'une espèce donnée de cette variation, sont importants dans la compréhension des forces évolutives. La plupart des études scientifiques sur la variation quantitative chez les plantes se sont intéressées à la morphologie et à la phénologie mais pas aux caractères impliqués dans le succès reproducteur. L'objectif de cette thèse est de mieux comprendre comment la répartition de la variation à des marqueurs neutres et des caractères quantitatifs influence l'évolution de la reproduction et des mécanismes de défense dans six populations Européennes de l'espèce dioïque Silene latifolia. Nous avons mis en évidence une grande diversité intra et inter-population à six loci microsatellites ainsi qu'à la plupart des caractères quantitatifs mesurés (morphologie, phénologie et traits reproducteurs) sur des plantes cultivées dans des conditions standardisées. Un résultat intéressant est la présence d'un cline latitudinal pour l'âge à la floraison. Ceci est probablement une conséquence de la sélection naturelle, puisque ce caractère est différencié entre les populations étudiées, héritable et que la différenciation de ce trait est supérieure à la différenciation des marqueurs neutres étudiés. Notre étude a ensuite porté plus précisément sur les pressions de sélection spécifiques aux sexes : la coadaptation entre les sexes et les mécanismes de défense contre l'insecte granivore Hadena bicruris. Afin d'évaluer la divergence sur les traits reproducteurs, nous avons étudié les effets des populations d'origine des mâles et des femelles et en particulier le succès reproducteur des mâles après pollinisation à l'aide de croisements inter et intra-population. Nous avons pollinisé la même femelle avec du pollen provenant d'un mâle de la même population ainsi qu'avec le pollen de deux mâles provenant de deux autres populations en situation de compétition avec un pollen provenant d'une population test. Des croisements contrôle ont été réalisés avec les mêmes mâles en pollinisation pure. Nous avons évalué le succès reproducteur de chaque mâle à l'aide d'analyses de paternité ainsi que la production de graines et la fitness de la descendance. L'origine du mâle avait un effet sur la paternité. Le taux de croissance in vitro du pollen est un caractère héritable et a eu un effet positif sur le succès reproducteur. De plus, l'origine de la femelle avait un effet sur le succès des mâles en compétition dans certaines populations. Nos résultats suggèrent qu'une divergence reproductive chez Silene latifolia pourrait apparaître suite à la compétition pollinique. Nous avons ensuite testé des mécanismes potentiels de défense constitutive et induite contre l'herbivore spécialiste Hadena bicruris, un papillon nocturne qui pourrait aussi jouer un rôle dans la différenciation des populations. L'épaisseur des fruits étant plus faible dans les régions où la plante est invasive (Amérique du Nord) et où l'insecte est absent, ce trait pourrait jouer un rôle défensif. Une pression de sélection plus faible causée par l'absence de l'herbivore aurait abouti à une perte de cette défense dans ces régions. Nous avons montré que l'épaisseur du fruit est variable selon les populations. L'infestation artificielle de fruit par l'insecte induit l'abscission sélective des fruits. Les fruits produits après une infestation n'étaient pas plus épais que les fruits issus de plantes non infestées. De plus, les fruits épais n'étaient pas moins nutritifs et ne causaient pas de perte de temps pour la prédation pour les larves, ce qui suggère que l'épaisseur des fruits ne joue pas un rôle défensif. Nos résultats montrent que plusieurs pressions de sélection interviennent et interagissent dans l'évolution de populations distantes, provoquant la divergence des populations Européennes de l'espèce Silene latifolia.

Diskreetin optimoinnin käyttö mekatronisen koneen virtuaalisuunnittelussa

Työn tavoitteena oli selvittää konetekniikan osastolla valmistetun optimointiohjelman soveltuvuutta virtuaaliprototyyppisen optimointiin. Lisäksi työn tavoitteena oli selvittää virtuaaliprototyyppien avulla tapahtuvan optimoinnin rajoitteet ja mahdollisuudet todellisilla optimointitehtävillä. Optimaze-ohjelma yhdistettiin simulointiohjelmistoon käyttäen apuna merkkitiedostoja ja simulointiohjelmiston sisäisiä makroja. Saadun optimointiympäristön toimivuus testattiin kahdella todellista puomia optimoivalla optimointitehtävällä. Simulointiohjelmistona käytettiin ADAMS:ia ja optimointialgoritmina differentiaalievoluutiota. Tuloksista havaittiin optimointiohjelman soveltuvan virtuaaliprototyyppien optimointiin. Raskaiden mallien optimoinnin huomattiin kuitenkin olevan liian hidas prosessi. Tutkimuksessa todettiinkin asian vaativan lisää tutkimista ja kehitystyötä.

Predicting diffusion of innovations with self-organisation and machine learning

The main subject of this master's thesis was predicting diffusion of innovations. The prediction was done in a special case: product has been available in some countries, and based on its diffusion in those countries the prediction is done for other countries. The prediction was based on finding similar countries with Self-Organizing Map~(SOM), using parameters of countries. Parameters included various economical and social key figures. SOM was optimised for different products using two different methods: (a) by adding diffusion information of products to the country parameters, and (b) by weighting the country parameters based on their importance for the diffusion of different products. A novel method using Differential Evolution (DE) was developed to solve the latter, highly non-linear optimisation problem. Results were fairly good. The prediction method seems to be on a solid theoretical foundation. The results based on country data were good. Instead, optimisation for different products did not generally offer clear benefit, but in some cases the improvement was clearly noticeable. The weights found for the parameters of the countries with the developed SOM optimisation method were interesting, and most of them could be explained by properties of the products.

Many valued algebraic structures as measures of comparison

This thesis studies the properties and usability of operators called t-norms, t-conorms, uninorms, as well as many valued implications and equivalences. Into these operators, weights and a generalized mean are embedded for aggregation, and they are used for comparison tasks and for this reason they are referred to as comparison measures. The thesis illustrates how these operators can be weighted with a differential evolution and aggregated with a generalized mean, and the kinds of measures of comparison that can be achieved from this procedure. New operators suitable for comparison measures are suggested. These operators are combination measures based on the use of t-norms and t-conorms, the generalized 3_-uninorm and pseudo equivalence measures based on S-type implications. The empirical part of this thesis demonstrates how these new comparison measures work in the field of classification, for example, in the classification of medical data. The second application area is from the field of sports medicine and it represents an expert system for defining an athlete's aerobic and anaerobic thresholds. The core of this thesis offers definitions for comparison measures and illustrates that there is no actual difference in the results achieved in comparison tasks, by the use of comparison measures based on distance, versus comparison measures based on many valued logical structures. The approach has been highly practical in this thesis and all usage of the measures has been validated mainly by practical testing. In general, many different types of operators suitable for comparison tasks have been presented in fuzzy logic literature and there has been little or no experimental work with these operators.

Minimum-uncertaity states and pseudoclassical dynamics. II

The set of initial conditions for which the pseudoclassical evolution algorithm (and minimality conservation) is verified for Hamiltonians of degrees N (N>2) is explicitly determined through a class of restrictions for the corresponding classical trajectories, and it is proved to be at most denumerable. Thus these algorithms are verified if and only if the system is quadratic except for a set of measure zero. The possibility of time-dependent a-equivalence classes is studied and its physical interpretation is presented. The implied equivalence of the pseudoclassical and Ehrenfest algorithms and their relationship with minimality conservation is discussed in detail. Also, the explicit derivation of the general unitary operator which linearly transforms minimum-uncertainty states leads to the derivation, among others, of operators with a general geometrical interpretation in phase space, such as rotations (parity, Fourier).

Risk Parity: A Modern Tool for Diversification

The purpose of this study is to examine how well risk parity works in terms of risk, return and diversification relative to more traditional minimum variance, 1/N and 60/40 portfolios. Risk parity portfolios were constituted of five risk sources; three common asset classes and two alternative beta investment strategies. The three common asset classes were equities, bonds and commodities, and the alternative beta investment strategies were carry trade and trend following. Risk parity portfolios were constructed using five different risk measures of which four were tail risk measures. The risk measures were standard deviation, Value-at-Risk, Expected Shortfall, modified Value-at-Risk and modified Expected Shortfall. We studied also how sensitive risk parity is to the choice of risk measure. The hypothesis is that risk parity portfolios provide better return with the same amount of risk and are better diversified than the benchmark portfolios. We used two data sets, monthly and weekly data. The monthly data was from the years 1989-2011 and the weekly data was from the years 2000-2011. Empirical studies showed that risk parity portfolios provide better diversification since the diversification is made at the risk level. Risk based portfolios provided superior return compared to the asset based portfolios. Using tail risk measures in risk parity portfolios do not necessarily provide better hedge from tail events than standard deviation.