This thesis is aimed to empirically investigate the performance impact of frequent acquisitions as an aggressive merger and acquisition (M&A) strategy for an acquiring firm. In literature related to the study of M&A, a common question is whether acquisitions improve the performance of acquirers. Neither theoretical nor empirical studies have a clear view on the performance effect of M&A. Some argue positively and some are opposite. Although existing research are mixed for their arguments, a takeover is commonly perceived as a shock to the firm with a constant effect on changing business performance. This static perception of M&A creates a difficulty in explaining why firms acquire others when the performance effect is negative. To address the issue, this thesis examines the M&A effect dynamically with taking into account the role of merger frequency in affecting performance. On the basis of a large sample that consists of about 14,000 acquisitions from more than 100 countries over last 12 years, the thesis finds that the investors perceive a lower value if the acquiring firm is involved in frequent mergers. This is because more mergers are expected to attract considerable amount of management attention away from profitable activities in order to digest the challenges of new business integration at least in the short run. This “digesting constraint” argument is evident by our estimations. Firm becomes less profitable in the short run after a merger shock, and this adverse effect can be more severe if the firm is involved in more frequent mergers. Evidence of the thesis further show that, the effect of merger shocks is not static and persistent, and it changes with time. The shock affects adversely profitability in the short run, usually lasting a couple of years, and then the negative effect on performance could be turned either oppositely if the firm digests the shock successfully, or otherwise, continuously but diminishing over time if the digestion takes longer such as for frequent acquisition. This finding implies that the pace of firm resilience to a merger shock can be affected by its merger strategies. The pace can be slow if the firm pursues frequent mergers aggressively. The performance effect of a merger shock is dynamic and changes with time. The dynamic view for merger shocks from this study opens a new vision for literature in merger studies. Overall the market expectation to a merger effect on changing firm performance is quite consistently related to what has actually happened to the firm after the merger shock.

[pt] Este trabalho apresenta uma análise do processo de fusão de uma substância pura em uma cavidade retangular; a qual tem uma das paredes verticais aquecida. A fase sólida do material em mudança de fase é mantida isotérmica à temperatura de fusão. Consequentemente, todo o processo é controlado pela convecção natural térmica na fase líquida. Os gradientes de densidade são devidos unicamente ao campo de temperaturas. O escoamento é laminar e a análise é feita de forma bidimensional. O fluido de trabalho é newtoniano. Supõe-se que a hipótese de Boussinesq é válida, ou seja, considera-se que o fluido possui densidade constante, com exceção do termo de força de empuxo da equação de momentum. São propostas algumas aproximações para descrever o comportamento da densidade em função da temperatura neste termo. As aproximações propostas dependem da substância de trabalho e da gama de temperaturas considerada. Materiais com diferentes números de Prandtl, sujeitos à fusão são analisados numericamente. A simulação numérica utiliza o Método dos Volumes Finitos e é baseada em uma hipótese que considera o processo de fusão como uma sucessão de estados quase-estacionários. Assim, a partir de uma dada posição da frente de fusão, resolve-se as equações de convecção natural em uma cavidade fixa. O formato irregular da interface sólido-líquido é levado em conta através de uma transformação de coordenadas não-ortogonais. De posse da solução da transferência de calor na interface, pode-se determinar a posição da frente de fusão no final do passo de tempo considerado. Em seguida, retorna-se o cálculo do escoamento convectivo para a nova cavidade líquida. O processo é repetido até que o processo de fusão chegue próximo ao fim. O caso da água, que passa por um máximo de densidade próximo à temperatura de 4 graus Celsius, recebe atenção especial. O comportamento anômalo da densidade desta substância pode exercer forte influência sobre a transferência de calor durante o processo de fusão do gelo. O interesse particular no caso da água se deve ao fato do máximo de densidade ocorrer em um domínio de temperaturas frequentemente encontrado na natureza e em diversas aplicações tecnológicas. A carência de trabalhos neste campo motivou a realização de uma parte experimental. Os experimentos realizados permitem medidas de temperatura e observações fotográficas do escoamento e do avanço da frente de fusão. Os resultados mostram que a variação não-linear da densidade do termo de força de empuxo afeta a transferência de calor por convecção natural, o movimento da interface e a estrutura do escoamento. / [en] This work presents an analysis of the melting process of a pure substance enclosed in a rectangular cavity heated from one of the verticals walls. The solid phase of the phase-change material is maintained isothermal at the fusion temperature. Consequently, the process is controlled by thermal natural convection in the liquid phase. The density gradients result from the temperature field only. The flow is laminar and the analysis is two-dimensional. The fluid is Newtonian. The Boussinesq hypothesis is assumed, that is the fluid density is considered to be constant, except in the driving term of the momentum equation. Various approximations are proposed for this term to describe the density behavior with the temperature. The proposed approximations depend on the substance and on the temperature range. Materials submitted to the melting phenomena with different Prandtl numbers are numerically investigated using a Finite Volume Method. The numerical simulation is based on the hypothesis of the quasi-stationary processes. So, from a certain position of the melting front, the natural convection equations in a fixed cavity are solved. The shape of the solid-liquid interface creates some difficulties. An algebraic method of transformation of non-orthogonal coordinates is used to map the irregular physical space occupied by the liquid into a rectangular computational space. The resulting local heat transfer at the interface is then used to determine the position of the melting front at the end of the time step. Next, the convective flow at the new liquid cavity is calculated. The procedure is repeated until the substance is almost completely melted. Special attention is given to the case of water, where a maximum density is attained at a temperature close to 4 degrees Celsius. The anomalous behavior of the water density can strongly influence the heat transfer during the melting process of ice. The lack of works in this field have motivated an experimental investigation. The particular interest of ice melting is based on the fact that the maximum density occurs over a wide range of temperature frequently found in nature and many technological applications. The experiments performed allow temperature measurements and photographic observations of the flow and the melting front motion. The results show that the non-linear density variation in the buoyancy term affects the heat transfer by natural convection, the interface motion and the flow structure.