Abstract Recently, domestic mutual fund development was rapidly. As widely open door for foreign investment and mutual fund joining stock market, its structure starts slowly changing. Via different kinds of expansion and merger let The properties of companies tend to complex. Investors sometimes hardly make correction decision. Here just focus on mutual funds which are belong to common stock type to observe the existence of herding trade. And then check the relation between performance roll drift and herding trade in mutual funds. Usually known how to select funds is performance roll. And general one likes to choose No. 1 which may be myth. Actually No. 1 fund sometimes put heavy weight on particular stock. So it might temporary catch head. But it does not gain steady return. Then selecting steady head group is more important. Is it true? Investment companies may issue several funds. During inside discuss between funds managers could influence stock holding. Even though information spread speed could have different between crowd and investment com- panies, they are all affected by outside. So call hedging is mean that follows market behavior to cow buy or bear sell. If the choice of funds holding stocks follows other funds, perform- ance roll will random shift. Then using it to judge funds selection would not be reliable. The major are common stock type mutual funds belong to Taiwan mutual fund perform- ance. Draw funds front 20 and rear 20 and define them as winner and loser. Check their stock ownerships correlation, herding coefficient and drift phenomenon in certain period. During time set, find winner and loser have preference to keep some group stocks. In co- mparsion with entire common stocks funds, herding values in winner and loser are not so obviously different with zero. Loser group drift to winner side when values are more different with zero. In certain period, there is high drift ration which change performance roll list. Keyword: hedging¡Bfund performance¡Bwinner¡Bloser

<p>In this thesis we consider two models for the computation of option prices. The first one is a generalization of the Black-Scholes model. In this generalization the volatility Sigma is not a constant. In the simplest case it changes at once at a certain time moment Tau. In some sense this is the conditionally Levy model. For this generalized Black-Scholes model have been theoretically obtained formulas for vanilla Call/Put option prices. Under the assumption of a good prediction of the parameter Sigma the obtained numerical results fit the real dara better than standard Black-Scholes model.</p><p>Second model is an exponential Levy model, where a Levy process is the CGMY process. We use the finite-difference scheme for computations of option prices. As example we consider vanilla Call/Put, Double-Barrier and Up-and-out options. After the estimation of the parameters of the CGMY process by the method of moments we obtain options prices and calculate fitting error. This fitting error for the CGMY model is smaller than for the Black-Scholes model.</p>

In this thesis we consider two models for the computation of option prices. The first one is a generalization of the Black-Scholes model. In this generalization the volatility Sigma is not a constant. In the simplest case it changes at once at a certain time moment Tau. In some sense this is the conditionally Levy model. For this generalized Black-Scholes model have been theoretically obtained formulas for vanilla Call/Put option prices. Under the assumption of a good prediction of the parameter Sigma the obtained numerical results fit the real dara better than standard Black-Scholes model. Second model is an exponential Levy model, where a Levy process is the CGMY process. We use the finite-difference scheme for computations of option prices. As example we consider vanilla Call/Put, Double-Barrier and Up-and-out options. After the estimation of the parameters of the CGMY process by the method of moments we obtain options prices and calculate fitting error. This fitting error for the CGMY model is smaller than for the Black-Scholes model.

Animals that live in a social group are often organized in a hierarchy with rank determining access to resources. Maintaining a dominant position requires a high rate of energetically expensive aggressive displays and physical exertion. Lab based winnerloser studies, often conducted with individuals from non-social species, have shown that subordinates have higher stress hormone levels than dominant individuals (subordinatestress hypothesis). However, in carnivorous animals that are cooperative breeders, displays of aggression are associated with elevated stress hormone levels (dominancestress hypothesis). The effect of reproductive state on dominance and stress is not addressed within either of these hypotheses. The purpose of this study was to examine stress level in relation to dominance rank and reproductive state in a non-cooperative breeder and herbivore, the domestic horse. As rank and reproductive state can affect behavior, I examined activity budgets, behavioral patterns, and social interactions, as well as the proximity and identity of neighbors in the social group. At two facilities in Kentucky, I recorded the social interactions and measured fecal glucocorticoids in pastured, female horses that were either lactating or non-lactating. While fecal glucocorticoid level did not differ between reproductive state and rank, activity behavior demonstrated significant differences between reproductive states. Higher energetic requirements of lactation were reflected in significantly longer bouts of eating and significantly less time spent alert and socializing. The non-cooperative social nature of horses does not limit their reproduction or resource acquisition based upon rank, and therefore does not fit with the dominance-stress hypothesis or subordinate-stress hypothesis and instead offers the alternative of an independent-stress hypothesi

A key problem in designing marketplaces is how to efficiently allocate a collection of goods among multiple people. Auctions have emerged as a powerful tool with the promise to increase market efficiency by allocating goods to those who value them the most. Nevertheless, traditional auctions are unable to handle real-world market complexities. Over the past decade, there has been a trend towards allowing for package bids and other types of multidimensional bidding techniques that enable suppliers to take advantage of their unique abilities and put forth their best offers. In particular the application of iterative combinatorial auctions in procurement saves negotiation costs and time. Conceptually these auctions show a potential for improving the overall market efficiency. However, in practice they host several new challenges and difficulties. One challenge facing the auctioneer in an iterative combinatorial auction environment is to quickly find an acceptable solution for each round of the auction. Bidders require time to precisely evaluate, price, and communicate different possible combinations based on their current information of item prices. The auctioneer requires time to solve the underlying mathematical problem formulation based on the bids received, report back the feedback information and initiate a new round of the auction. In Chapter 3, we propose a Lagrangian-based heuristic to solve the auctioneer's winner determination problem. After generating the Lagrange multipliers from the solution of a linear relaxation, the heuristic applies several procedures to fix any potentially infeasible optimal Lagrange solutions. In addition to providing an efficient way of solving the winner determination problem, as compared with the leading commercial solver CPLEX, our approach provides Lagrange multipliers. The latter are used as proxies for prices in the auction feedback mechanism. In Chapter 4 we develop a model for the bidders pricing problem, an issue that has received much less attention in the literature. Using the auctioneer feedback, that includes the Lagrange multipliers, the pricing model maximizes the bidders' profit while at the same time keeping their bids competitive. We derive several optimality results for the underlying optimization problem. Interestingly, we analytically show that the auction converges to a point where no bidder is able to submit a bid that yields strictly better profit for him and is not less competitive than his previous bids submitted. We experimentally observe that this approach converges in an early stage. We also find that this iterative auction allows the bidders to improve their profit while providing lower and competitive prices to the auctioneer. In Chapter 5, we introduce a flexible auction model that allows for partial bids. Rather than the regular all-or-nothing indivisible package bids, divisible bids provide flexibility for the auctioneer with the possibility to accept parts of the bids and yet allow the suppliers to capture synergies among the items and provide quantity discounts. We show numerically that this approach improves the overall efficiency of the auction by increasing the suppliers' profit while decreasing the auctioneer's total price of procurement. In addition, we find that computationally the flexible auction outperforms the regular auction. / Thesis / Doctor of Philosophy (PhD)

The Winner-Take-All (WTA) is an essential neural network operation for locating the most active neuron. Such a procedure has been extensively used in larger application areas. The Winner-Take-All circuit selects the maximum of the inputs inhibiting all other nodes. The efficiency of the analog circuits may well be considerably higher than the digital circuits. Also, analog circuits’ design footprint and processing time can be significantly small. A current-based Winner-Take-All circuit for analog neural networks is presented in this research. A compare and pass (CAP) mechanism has been used, where each input pair is compared, and the winner is selected and passed to another level. The inputs are compared by a sense amplifier which generates high and low voltage signals at the output node. The voltage signal of the sense amplifier is used to select the winner and passed to another level using logic gates. Also, each winner follows a sequence of digital bits to be selected. The findings of the SPICE simulation are also presented. The simulation results on the MNIST, Fashion-MNIST, and CIFAR10 datasets for the memristive deep neural network model show the significantly accurate result of the winner class with an average difference of input and selected winner output current of 0.00795uA, 0.01076uA and 0.02364uA respectively. The experimental result with transient noise analysis is also presented.

This dissertation investigates the long-term performance of momentum portfolios. Its results show striking asymmetries for winners and losers and imply potentially different causes for the winner and loser components of momentum. After separately examining winners and losers relative to their respective benchmark portfolios with no momentum, we find winner momentum is smaller in magnitude, persists only for six months, and its higher return fully reverses. This is consistent with the notion that winner momentum is an overreaction to positive news and potentially destabilizing. Loser momentum is larger in magnitude, lasts for about one year, and its lower return does not reverse in the long run. This is consistent with the notion that loser momentum is an underreaction to negative news and suggests investors hold on to losers for too long. The lack of reversal for losers departs from prior studies whose findings are driven by the use of monthly rebalanced portfolio. Rebalancing cumulates an upward bias caused by noise-induced price volatility, which disproportionately affects losers more. This greater upward bias in losers creates an illusion that the winner minus loser return reverses. More appropriate approaches such as the buy-and-hold portfolio documents significantly less reversal. Existing theories that potentially conform to the overreaction of winners and underreaction of losers include overconfidence (Daniel, Hirshleifer, and Subrahmanyam, 1998), representativeness and conservatism (Barberis, Shleifer, and Vishny, 1998), interaction between agents holding asymmetric information (Hong and Stein, 1999), and investors' asymmetric response to fund performance (Vayanos and Woolley, 2013). / Doctor of Philosophy / The method employed to study a phenomenon can have an immense impact on our understanding. In the specific context of momentum - a strategy of buying stocks with good past performance and selling stocks with bad past performance, we show that the methodology choices as simple as how you form a portfolio and what you benchmark the portfolio against can produce significantly different results than previously documented. The documented pattern that momentum reverses over the long run, is confounded by the use of rebalanced portfolio and benchmarking winner and loser stocks against each other. Rebalancing embeds an upward bias into the return, with the bias increasing in the amount of noise in the price. Losers, having lower prices and smaller market values, suffer more from the upward bias. Thus, the reversal of the winner over loser return is more due to the inflated loser return from the bias than an underlying economic phenomenon. We confirm this by using a host of other portfolio formation methods that are known to mitigate the bias. With each of the other method, the loser return and the reversal are significantly reduced. We also suggest comparing winners and losers to a neutral benchmark with no momentum, rather than with each other to study the long-term pattern of momentum. This exercise turns out to be fruitful in that we find asymmetric behavior for winners and losers. Winner momentum fully reverses while loser momentum persists. The significance of the new results is that they affect our understanding of what drives momentum in the first place. The reversal of winners implies that winner momentum is an overreaction to positive news and potentially destabilizing. The persistence of losers indicates that loser momentum is an underreaction to negative news and implies investors hold on to losers for too long.

L'exploitation, le contrôle et la signalisation des systèmes de métros modernes et en particulier les métros sans conducteur, reposent sur deux familles de systèmes de communication radio sans fil : les transmissions vitales pour la signalisation et le contrôle des trains à faible débit et les transmissions non vitales très haut débit pour la vidéo surveillance embarquée, le télé-diagnostic,l'information des clients, etc. Lors de la mise en œuvre de tels systèmes de transmission, l'industriel doit donc garantir à l'exploitant de métro des performances concernant les débits et la qualité de la transmission en termes de perte de paquets d'information. Dès lors, appréhender le comportement du canal de propagation est un élément clé pour prédire et améliorer les performances des systèmes de transmission.L'objectif de ces travaux de thèse est double et concerne :• D'une part, la réduction des temps de calcul d'un simulateur de canal à tracé de rayons 3D qui prédit avec précision le comportement du canal de propagation, mais reste coûteux en temps de calculs pour les simulations dynamiques. Nous proposons dans cette thèse une méthode qui s'appuie sur trois critères de visibilité pour simplifier la description de lagéométrie de l'environnement sans pénaliser la prédiction des paramètres caractéristiques du canal de propagation ;• D'autre part, la mise en œuvre d'un nouveau modèle de canal de propagation MIMO dynamique en tunnel rectiligne de section rectangulaire qui permettra d'optimiser des systèmes de transmission multi antennaires (MIMO) pour des applications de transmissions sans fil pour les métros en tunnel. Ce modèle s'inspire globalement du modèle de canal utilisé / The exploitation, control and signalling systems for metro and especially modern driverless metros, are based on two families of radio communication systems Wireless : vital transmissions for signalling and train control low-flow and nonvital transmissions with very high bandwidth for video-tracking systems, telediagnosis, customer information, etc.. During the implementation ofsuch transmission systems, the industrial must therefore ensure the metro operator performance on throughput and transmission quality in terms of packet loss information. Therefore, understanding the behavior of the propagation channel is a key to predict and improve performance of transmission systems.The aim of this thesis is twin and concerns :• On the one hand, the computation time reduction of a 3D ray tracing simulator that accurately predicts the behavior of the propagation channel, but it is expensive in times of calculation for dynamic simulations. We propose here a new method based on three visibility criteria to simplify the geometry description of the environment without degrading the characteristic parameters prediction of the propagation channel ;• On the other hand, the construction of a new model for MIMO propagation channel dynamics in straight tunnel of rectangular section which will optimize transmission systems multi-antennary (MIMO) applications for wireless transmission in metros tunnel. This model draws broadly from the channel model used in the standard WINNER and is fed by the results extracted from the 3D ray tracing channel simulator. The results obtained in this thesis are encouraging and offer many opportunities.

A seleção de fornecedores de transporte é um desafio cada vez maior. O crescimento da rede de clientes a ser coberta demanda uma alocação eficiente em termos de custo não suprida por mecanismos tradicionais de negociação. Neste âmbito, o leilão combinatório torna-se uma alternativa de negociação ao permitir capturar sinergias entre os trajetos que devem ser atendidos. Em conseqüência disso, diminui-se o custo de transporte do fornecedor que se reflete nos menores preços de suas propostas e finalmente no custo total de compra do serviço. Por outro lado, esta decisão envolve fatores além do custo total; a mensuração destes torna-se importante para identificar fornecedores que melhor se ajustam aos requerimentos do comprador. No entanto, é fundamental escolher um método adequado para sua avaliação porque este influência a decisão final. Este problema de compra de serviços de transporte é conhecido na literatura como Winner Determination Problem (WDP) que, devido a sua complexidade, possui uma resolução limitada. Após revisão teórica, foi observado que os estudos relacionados à área de transporte focalizavam o desenvolvimento de modelos matemáticos que fossem representativos da realidade. Alguns destes modelos abordam a utilização de múltiplos critérios atribuindo um coeficiente que pondera cada critério. Evidenciou-se a necessidade do desenvolvimento de um algoritmo alternativo que além de facilitar sinergias entre trajetos, fosse abrangente o suficiente para tratar múltiplos critérios em instâncias compatíveis com problemas reais. Logo, com o intuito de contribuir com a literatura foi adaptado um algoritmo matemático otimizante ao problema de compras de fornecedores de transporte com base no algoritmo de Sandholm (2002). Este algoritmo aplica leilão combinatório de compras, apoiando-se na teoria da análise de decisão para mensurar critérios relevantes do comprador. Inicialmente, o algoritmo minimiza o custo total do comprador designando combinações de trajetos e fornecedores; depois é modificado para o tratamento multi-critério. Os resultados obtidos foram comparados com o software comercial CPLEX. / Selecting suppliers is a crescent challenge for the enterprises. The extent of the client web that needs to be served demands efficient allocations, in terms of cost, that are not addressed by traditional mechanisms. In this scenario, another mechanism came to be: the combinatorial auction. In this one, suppliers can express their synergies on routes they wish to supply. This leads to lowering their transportation costs, which is reflected in lower bidding prices as well as in the total cost of service. On the other hand, the selection of a supplier involves other criteria besides cost. The definition of these is essential to define which supplier fits the needs of the buyer. That is why it is of most importance to choose the right method to evaluate these needs, as it defines the final choice. This problem is known as Winner Determination Problem (WDP) and due to its complexity, possesses a feeble solution. After compiling what has been done about the subject, it was noticed that in the field of transport, studies are focused on mathematical models that represent reality. Some models address criteria assigning coefficients to the objective function by weighting on it. Clearly, there was a need for alternative algorithms that would, besides promoting synergies on routes, also treat multi-criteria problems close to reality. Therefore, searching for a valid contribution in the field, an adaption of an optimizing algorithm based on Sandholm (2002)s was made. The algorithm applies combinatorial auction, supported by decision analysis for measuring relevant buyers criteria. First, the main algorithms objective is to minimize buyers costs by combining routes and suppliers; then, a modified approach considers multi criteria. Results were then compared to the commercial software CPLEX.

La nécessité d'une transmission sans fils des données à des débits élevés, à la fois fiables et avec de faible latence a donné lieu à ces dernières années à une succession de normes sans fil, allant de 3G-4G, WLAN à la cinquième génération (5G) des réseaux mobiles. Dans ce contexte, les équipementiers, ainsi que les opérateurs, doivent élaborer des méthodes d'essai standard précises et efficaces pour évaluer les performances des systèmes et des terminaux. Les méthodologies de test en direct par voie aérienne ("Over-The-Air") (OTA) visent à reproduire des environnements multi-trajets radio en laboratoire de manière répétable et contrôlable, en évitant les coûteuses mesures in-situ. L'objectif de cette thèse est de proposer une nouvelle méthodologie d'essai OTA, afin de reproduire la propagation des canaux radio, sur une large bande et d'évaluer les performances des systèmes sans fil dans des environnements réels. La thèse débute en présentant les bases de la chaîne radio et de certains modèles de chaînes présentés dans la littérature. Ensuite, un examen critique des méthodologies OTA existantes dans la littérature est fourni. Parmi les différentes méthodologies, nous avons opté pour l'approche de la chambre anéchoïde multi-sonde, qui consiste à déployer un certain nombre de sondes autour d'un équipement radio sous test et à les alimenter avec un émulateur d’évanouissements (fading). Cette méthodologie fournit une reproduction précise des caractéristiques des canaux spatiaux, qui sont nécessaires pour évaluer la performance des terminaux multi-antennes dans des environnements réels. L'avantage le plus important de cette méthodologie est la capacité d'imiter différents modèles de canaux en termes de résolution spatiale, d’évanouissements angulaire et temporel. Un outil de simulation a été développé pour étudier et déterminer les caractéristiques de l'installation OTA pour différents types de canaux d’intérêt. En particulier, le nombre et la mise en place des antennes nécessaires et la taille de l'installation ont été étudiés en fonction de la taille électrique du dispositif testé. Sur la base des études de dimensionnement, une configuration OTA expérimentale a été réalisée pour reproduire les caractéristiques des canaux dans l'espace tridimensionnel pour une plage de fréquences de 2 à 6 GHz. / The need for high data-rate, reliable and low latency transmission in wireless communication systems motivated a multitude of wireless standards, spanning from 3G-4G, WLAN to the upcoming fifth generation (5G) of mobile networks. In this context, technology providers, as well as operators, need to develop accurate and cost effective standard test methods, to evaluate devices performance. Over-The-Air (OTA) test methodologies aim to reproduce radio multipath environments in laboratory in repeatable and controllable manner, avoiding costly field test. The focus of this thesis is to propose a new OTA test methodology, in order to emulate radio channel propagation, over a wide band, and to evaluate the performance of the wireless systems in real environments. We start our study by introducing the basics of radio channel and some channel models presented in literature. Then a critical review of existing OTA methodologies in literature is provided. Among the different methodologies we opted for the multi-probe anechoic chamber approach, which consists into deploying a number of probes around a device, and feed them with fading emulator. This methodology provides an accurate reproduction of spatial channel characteristics, which are needed to assess the performance of multi-antenna terminals in real environments. The most important advantage of this methodology is the capability to emulate different channel model in term of spatial resolution, angular and temporal fading. A simulation tool was developed to investigate and determine the OTA setup under different channel condition. In particular the number and emplacement of antennas needed and the size of the setup were investigated as a function of the electrical size of the device under test. Based on the dimensioning studies, an experimental OTA setup was realized to reproduce the channel characteristics in the three dimensional space for a frequency range from 2 to 6 GHz.