Capacidad de obtener beneficio de reglas técnicas de contratación basadas en algoritmos de minería de datos

Trabajo Fin de Grado de la doble titulación de Grado en Ingeniería Informática y Grado en Administración y Dirección de Empresas.

The social construction of cooperation: exchange, identification and collective intentions within the organization

My aim is to develop a theory of cooperation within the organization and empirically test it. Drawing upon social exchange theory, social identity theory, the idea of collective intentions, and social constructivism, the main assumption of my work implies that both cooperation and the organization itself are continually shaped and restructured by actions, judgments, and symbolic interpretations of the parties involved. Therefore, I propose that the decision to cooperate, expressed say as an intention to cooperate, reflects and depends on a three step social process shaped by the interpretations of the actors involved. The first step entails an instrumental evaluation of cooperation in terms of social exchange. In the second step, this “social calculus” is translated into cognitive, emotional and evaluative reactions directed toward the organization. Finally, once the identification process is completed and membership awareness is established, I propose that individuals will start to think largely in terms of “We” instead of “I”. Self-goals are redefined at the collective level, and the outcomes for self, others, and the organization become practically interchangeable. I decided to apply my theory to an important cooperative problem in management research: knowledge exchange within organizations. Hence, I conducted a quantitative survey among the members of the virtual community, “www.borse.it” (n=108). Within this community, members freely decide to exchange their knowledge about the stock market among themselves. Because of the confirmatory requirements and the structural complexity of the theory proposed (i.e., the proposal that instrumental evaluations will induce social identity and this in turn will causes collective intentions), I use Structural Equation Modeling to test all hypotheses in this dissertation. The empirical survey-based study found support for the theory of cooperation proposed in this dissertation. The findings suggest that an appropriate conceptualization of the decision to exchange knowledge is one where collective intentions depend proximally on social identity (i.e., cognitive identification, affective commitment, and evaluative engagement) with the organization, and this identity depends on instrumental evaluations of cooperators (i.e., perceived value of the knowledge received, assessment of past reciprocity, expected reciprocity, and expected social outcomes of the exchange). Furthermore, I find that social identity fully mediates the effects of instrumental motives on collective intentions.

Il conflitto fra trasparenza sui mercati finanziari ed esigenza di riservatezza della società

The thesis main topic is the conflict between disclosure in financial markets and the need for confidentiality of the firm. After a recognition of the major dynamics of information production and dissemination in the stock market, the analysis moves to the interactions between the information that a firm is tipically interested in keeping confidential, such as trade secrets or the data usually covered by patent protection, and the countervailing demand for disclosure arising from finacial markets. The analysis demonstrates that despite the seeming divergence between informational contents tipically disclosed to investors and information usually covered by intellectual property protection, the overlapping areas are nonetheless wide and the conflict between transparency in financial markets and the firm’s need for confidentiality arises frequently and sistematically. Indeed, the company’s disclosure policy is based on a continuous trade-off between the costs and the benefits related to the public dissemination of information. Such costs are mainly represented by the competitive harm caused by competitors’ access to sensitive data, while the benefits mainly refer to the lower cost of capital that the firm obtains as a consequence of more disclosure. Secrecy shields the value of costly produced information against third parties’ free riding and constitutes therefore a means to protect the firm’s incentives toward the production of new information and especially toward technological and business innovation. Excessively demanding standards of transparency in financial markets might hinder such set of incentives and thus jeopardize the dynamics of innovation production. Within Italian securities regulation, there are two sets of rules mostly relevant with respect to such an issue: the first one is the rule that mandates issuers to promptly disclose all price-sensitive information to the market on an ongoing basis; the second one is the duty to disclose in the prospectus all the information “necessary to enable investors to make an informed assessment” of the issuers’ financial and economic perspectives. Both rules impose high disclosure standards and have potentially unlimited scope. Yet, they have safe harbours aimed at protecting the issuer need for confidentiality. Despite the structural incompatibility between public dissemination of information and the firm’s need to keep certain data confidential, there are certain ways to convey information to the market while preserving at the same time the firm’s need for confidentality. Such means are insider trading and selective disclosure: both are based on mechanics whereby the process of price reaction to the new information takes place without any corresponding activity of public release of data. Therefore, they offer a solution to the conflict between disclosure and the need for confidentiality that enhances market efficiency and preserves at the same time the private set of incentives toward innovation.

Bayesian change-point analysis in linear regression model with scale mixtures of normal distributions

In this thesis, we consider Bayesian inference on the detection of variance change-point models with scale mixtures of normal (for short SMN) distributions. This class of distributions is symmetric and thick-tailed and includes as special cases: Gaussian, Student-t, contaminated normal, and slash distributions. The proposed models provide greater flexibility to analyze a lot of practical data, which often show heavy-tail and may not satisfy the normal assumption. As to the Bayesian analysis, we specify some prior distributions for the unknown parameters in the variance change-point models with the SMN distributions. Due to the complexity of the joint posterior distribution, we propose an efficient Gibbs-type with Metropolis- Hastings sampling algorithm for posterior Bayesian inference. Thereafter, following the idea of [1], we consider the problems of the single and multiple change-point detections. The performance of the proposed procedures is illustrated and analyzed by simulation studies. A real application to the closing price data of U.S. stock market has been analyzed for illustrative purposes.

Portfolio-optimization models for small investors

Since 2010, the client base of online-trading service providers has grown significantly. Such companies enable small investors to access the stock market at advantageous rates. Because small investors buy and sell stocks in moderate amounts, they should consider fixed transaction costs, integral transaction units, and dividends when selecting their portfolio. In this paper, we consider the small investor’s problem of investing capital in stocks in a way that maximizes the expected portfolio return and guarantees that the portfolio risk does not exceed a prescribed risk level. Portfolio-optimization models known from the literature are in general designed for institutional investors and do not consider the specific constraints of small investors. We therefore extend four well-known portfolio-optimization models to make them applicable for small investors. We consider one nonlinear model that uses variance as a risk measure and three linear models that use the mean absolute deviation from the portfolio return, the maximum loss, and the conditional value-at-risk as risk measures. We extend all models to consider piecewise-constant transaction costs, integral transaction units, and dividends. In an out-of-sample experiment based on Swiss stock-market data and the cost structure of the online-trading service provider Swissquote, we apply both the basic models and the extended models; the former represent the perspective of an institutional investor, and the latter the perspective of a small investor. The basic models compute portfolios that yield on average a slightly higher return than the portfolios computed with the extended models. However, all generated portfolios yield on average a higher return than the Swiss performance index. There are considerable differences between the four risk measures with respect to the mean realized portfolio return and the standard deviation of the realized portfolio return.

Heterogeneous Beliefs and the Demand for D&O Insurance by Listed Companies

This paper introduces a new rationale for the existence of “Directors’ and Officers’” (D&O) insurance. We use a model with volatile stock markets where shareholders design compensation schemes that incentivize managers to stimulate short-term increases in stock prices that do not maximize long run stock market value. We show that D&O insurance provides a convenient instrument for the initial shareholders of a company to take advantage of differences in beliefs between insiders and outsiders in capital markets. The empirical results support the idea that both the insurance coverage and the premium are higher in the presence of new shareholders and volatile markets. The results prove robust in various empirical model specifications.

Portfolio-optimization models for small investors

Since 2010, the client base of online-trading service providers has grown significantly. Such companies enable small investors to access the stock market at advantageous rates. Because small investors buy and sell stocks in moderate amounts, they should consider fixed transaction costs, integral transaction units, and dividends when selecting their portfolio. In this paper, we consider the small investor’s problem of investing capital in stocks in a way that maximizes the expected portfolio return and guarantees that the portfolio risk does not exceed a prescribed risk level. Portfolio-optimization models known from the literature are in general designed for institutional investors and do not consider the specific constraints of small investors. We therefore extend four well-known portfolio-optimization models to make them applicable for small investors. We consider one nonlinear model that uses variance as a risk measure and three linear models that use the mean absolute deviation from the portfolio return, the maximum loss, and the conditional value-at-risk as risk measures. We extend all models to consider piecewise-constant transaction costs, integral transaction units, and dividends. In an out-of-sample experiment based on Swiss stock-market data and the cost structure of the online-trading service provider Swissquote, we apply both the basic models and the extended models; the former represent the perspective of an institutional investor, and the latter the perspective of a small investor. The basic models compute portfolios that yield on average a slightly higher return than the portfolios computed with the extended models. However, all generated portfolios yield on average a higher return than the Swiss performance index. There are considerable differences between the four risk measures with respect to the mean realized portfolio return and the standard deviation of the realized portfolio return.

Asset Pricing with Heterogeneous Agents, Incomplete Markets and Trading Constraints

The consumption capital asset pricing model is the standard economic model used to capture stock market behavior. However, empirical tests have pointed out to its inability to account quantitatively for the high average rate of return and volatility of stocks over time for plausible parameter values. Recent research has suggested that the consumption of stockholders is more strongly correlated with the performance of the stock market than the consumption of non-stockholders. We model two types of agents, non-stockholders with standard preferences and stock holders with preferences that incorporate elements of the prospect theory developed by Kahneman and Tversky (1979). In addition to consumption, stockholders consider fluctuations in their financial wealth explicitly when making decisions. Data from the Panel Study of Income Dynamics are used to calibrate the labor income processes of the two types of agents. Each agent faces idiosyncratic shocks to his labor income as well as aggregate shocks to the per-share dividend but markets are incomplete and agents cannot hedge consumption risks completely. In addition, consumers face both borrowing and short-sale constraints. Our results show that in equilibrium, agents hold different portfolios. Our model is able to generate a time-varying risk premium of about 5.5% while maintaining a low risk free rate, thus suggesting a plausible explanation for the equity premium puzzle reported by Mehra and Prescott (1985).

アフリカにおける株式市場の発展とその経済効果

本稿では、アフリカにおいて発展しつつある株式市場が、アフリカの各国経済に与える影響を分析する。最初に、アフリカの株式市場が、世界と比較すれば金額では小さいものの、アフリカの中ですでに活用されており、国際投資家からも注目されているという現実を明らかにする。次に、Sugimoto, Matsuki and Yoshida［2014］を概説し、アフリカ主要7カ国の株式市場のリターン（収益率）は、2004年以降、グローバル市場の動きから最も影響を受けており、世界金融危機などの際には平時よりも大きなグローバルショックを受けていたことを示す。最後に、アフリカ16カ国における株式市場の域内相互依存度を、固定相関係数（CCC-GARCH）モデルを用いて比較し、2012年以降、各地域共通の証券取引所創設にむけて積極的に連携姿勢を示すアフリカ諸国の間で、株式市場の地域的な連動性が高まってきていることを確認する。

Análisis fundamental aplicado a la industria minera : un estudio multivariante

Este trabajo tiene por objeto aplicar los principios del Value Investing a veinticuatro empresas del sector minero y definir las claves para extrapolar, en base a un análisis fundamental, una calificación para cada una de las empresas. Con este fin, se ha realizado un estudio estadístico multivariante para comparar las correlaciones existentes entre cada ratio fundamental y su evolución en bolsa a uno, tres y cinco años vista. Para procesar los datos se han utilizado los programas MATLAB y EXCEL. Sobre ellos se ha planteado una Matriz de Correlaciones de Pearson y un estudio de dispersión por cruce de pares. El análisis demostró que es posible aplicar la metodología del Value Investing a empresas del sector minero con resultados positivos aunque, el ajuste de las correlaciones, sugiere utilizar series temporales más largas y un mayor número de empresas para ganar fiabilidad en el contraste de estas hipótesis. De los estudios realizados, se deduce que unos buenos fundamentales influyen, de manera notable, a la revalorización bursátil a 3 y 5 años destacando, además, que el ajuste es mejor cuanto mayor sea este tiempo. Abstract This study aims to apply the principles of Value Investing to twenty four mining companies and, based on this fundamental study, develop a rating to classify those companies. For this purpose, we have performed a multivariate statistical study to compare the correlations between each fundamental ratio and its stock revalorization for one, three and five years. MATLAB and EXCEL have been used to process data. The statistical methods used are Pearson Matrix of Correlations and a Cross Pairs Scattering Study. The analysis showed that it is possible to apply the methodology of Value Investing to mining companies, although, the adjustment of correlations suggests using longer time series and a larger amount of companies to test these hypothesis. From the studies performed, it follows that good fundamentals significantly influence the stock market value at 3 and 5 years, noting that, the larger the period under study, the better the fit.

A linked data approach to sentiment and emotion analysis of twitter in the financial domain

Sentiment analysis has recently gained popularity in the financial domain thanks to its capability to predict the stock market based on the wisdom of the crowds. Nevertheless, current sentiment indicators are still silos that cannot be combined to get better insight about the mood of different communities. In this article we propose a Linked Data approach for modelling sentiment and emotions about financial entities. We aim at integrating sentiment information from different communities or providers, and complements existing initiatives such as FIBO. The ap- proach has been validated in the semantic annotation of tweets of several stocks in the Spanish stock market, including its sentiment information.

Una aproximación a la predicción del valor de acciones en la bolsa de valores aplicando técnicas de Data Mining

La predicción del valor de las acciones en la bolsa de valores ha sido un tema importante en el campo de inversiones, que por varios años ha atraído tanto a académicos como a inversionistas. Esto supone que la información disponible en el pasado de la compañía que cotiza en bolsa tiene alguna implicación en el futuro del valor de la misma. Este trabajo está enfocado en ayudar a un persona u organismo que decida invertir en la bolsa de valores a través de gestión de compra o venta de acciones de una compañía a tomar decisiones respecto al tiempo de comprar o vender basado en el conocimiento obtenido de los valores históricos de las acciones de una compañía en la bolsa de valores. Esta decisión será inferida a partir de un modelo de regresión múltiple que es una de las técnicas de datamining. Para llevar conseguir esto se emplea una metodología conocida como CRISP-DM aplicada a los datos históricos de la compañía con mayor valor actual del NASDAQ.---ABSTRACT---The prediction of the value of shares in the stock market has been a major issue in the field of investments, which for several years has attracted both academics and investors. This means that the information available in the company last traded have any involvement in the future of the value of it. This work is focused on helping an investor decides to invest in the stock market through management buy or sell shares of a company to make decisions with respect to time to buy or sell based on the knowledge gained from the historic values of the shares of a company in the stock market. This decision will be inferred from a multiple regression model which is one of the techniques of data mining. To get this out a methodology known as CRISP-DM applied to historical data of the company with the highest current value of NASDAQ is used.

Efectividad de las señales de compromiso de suscripción ante el anuncio de ampliaciones de capital en España

El objetivo de esta tesis doctoral es averiguar si el anuncio por parte del accionista significativo de ejercitar su derecho de suscripción preferente elimina o reduce la asimetría de información en las ampliaciones de capital con derecho de suscripción preferente en el Mercado de Valores español. Durante los 17 años analizados, encontramos que ni el anuncio de la ampliación de capital ni el tipo de aseguramiento acordado en cada ampliación tienen un impacto estadísticamente significativo en el Exceso de Rentabilidad Ajustada por Riesgo. Principalmente, el análisis realizado utiliza la información requerida por la Comisión Nacional del Mercado de Valores (CNMV) que deben aportar los accionistas significativos en el Folleto de emisión publicado con carácter previo a la ampliación. Esta investigación desglosa las ofertas en un Grupo 1, el cual incluye aquéllas en las que los accionistas significativos anuncian su intención de ejercitar su derecho en las emisiones, y el Grupo 2, que incluye aquéllas en donde no acuden o simplemente no existía información al respecto ya que no es una información obligatoria a incluir en el Folleto. Para cada ampliación de capital y para tres periodos de tiempo distintos se obtiene el Exceso de Rentabilidad Ajustada por Riesgo (ERAR) como la diferencia entre la Tasa Interna de Retorno y el Retorno Esperado, utilizando el modelo CAPM. De este modo, se trata de aislar el efecto temporal. La principal contribución de esta tesis doctoral es el hallazgo de una rentabilidad negativa estadísticamente significativa cuando el accionista significativo anuncia su intención de no suscribir la ampliación, o no existe información suficiente sobre su intención a este respecto. Adicionalmente, el análisis que se ha llevado a cabo en este estudio muestra un refuerzo estadísticamente significativo de este efecto negativo en la rentabilidad cuando existe simultáneamente una falta de compromiso por parte del accionista significativo y la ampliación no está asegurada. ABSTRACT The aim of this doctoral dissertation is to find out whether or not consideration of significant shareholders announcement of intention to exercise subscription rights makes a difference in eliminating or reducing the effects of asymmetrical information in equity offerings with pre-emptive rights on the Spanish Stock Market. For the 17 years of equity issues covered, we find that neither equity issue announcements nor the type of underwriting arrangements has a statistically significant impact on the issues’ Excess Risk Adjusted Return. The analysis uses the information required by CNMV (Spanish equivalent to SEC) to be provided by the significant shareholders in the equity issue’s prospectus. The doctoral dissertation breaks the offerings down into Group 1, in which the significant shareholders indicated their intention to subscribe, and Group 2, for which there was not enough information provided as to their intentions. For each equity issue, Excess Risk Adjusted Return (ERAR) is obtained, for three different periods, as is the difference between nominal Internal Rate of Return and expected return, using the CAPM. By subtracting the expected return from the IRR, the effect of time or any other variable influencing the stock price during the period, aside from the equity issue, should, in principle, be removed. The main contribution of this study is the finding of a statistically significant negative impact on returns either when the significant shareholders indicate their intention not to subscribe, or when not enough information is provided about their intention. We also find a statistically significant reinforcing negative effect on returns in the case of simultaneous lack of commitment on the part of significant shareholders, and non-underwritten equity issues.

O Conflito de Agência Exposto ao Efeito de Mercado: Um Estudo Empírico de Precificação de Ações da Bombril S/A

Esta dissertação apresenta um estudo exploratório que tem como objetivo a avaliação da reação do mercado frente aos problemas de agência e assimetria informacional entre os acionistas majoritários-controladores e os acionistas minoritários de uma empresa brasileira de capital aberto, no tocante ao valor das suas ações negociadas na Bolsa de Valores de São Paulo. O estudo se propõe a investigar o impacto de precificação dessas ações promovendo a análise do seu comportamento frente à divulgação de notícias de crise e fraude da empresa por meio da mídia especializada. A metodologia adotada consiste na aplicação de estudo de eventos para identificação de retornos anormais da empresa desde a divulgação da primeira notícia selecionada, datada de maio de 2001, até a última notícia em outubro de 2005, utilizando-se de procedimentos estatísticos como a regressão linear e aplicação do teste t de student para estimar e comparar os resultados. Os dados foram obtidos por meio do banco de dados da Economática Ltda, conforme acesso realizado na Universidade de São Paulo. Para objeto de pesquisa foi selecionada a empresa Bombril S/A, por estar em evidências quanto a problemas de agência no âmbito do mercado nacional. Os resultados obtidos apontaram que o mercado reagiu significantemente aos anúncios dos conflitos selecionados, apresentando um valor de p-value <0,05 para os blocos de eventos, o que significa a rejeição da hipótese nula, constatando que a evidencia estatística dos dados testados comprova retornos anormais acumulados diferentes de zero. Entretanto, sugere-se novas pesquisas com outros parâmetros de eventos na busca de mais evidências sobre o efeito das informações no preço das ações.

O Conflito de Agência Exposto ao Efeito de Mercado: Um Estudo Empírico de Precificação de Ações da Bombril S/A

Esta dissertação apresenta um estudo exploratório que tem como objetivo a avaliação da reação do mercado frente aos problemas de agência e assimetria informacional entre os acionistas majoritários-controladores e os acionistas minoritários de uma empresa brasileira de capital aberto, no tocante ao valor das suas ações negociadas na Bolsa de Valores de São Paulo. O estudo se propõe a investigar o impacto de precificação dessas ações promovendo a análise do seu comportamento frente à divulgação de notícias de crise e fraude da empresa por meio da mídia especializada. A metodologia adotada consiste na aplicação de estudo de eventos para identificação de retornos anormais da empresa desde a divulgação da primeira notícia selecionada, datada de maio de 2001, até a última notícia em outubro de 2005, utilizando-se de procedimentos estatísticos como a regressão linear e aplicação do teste t de student para estimar e comparar os resultados. Os dados foram obtidos por meio do banco de dados da Economática Ltda, conforme acesso realizado na Universidade de São Paulo. Para objeto de pesquisa foi selecionada a empresa Bombril S/A, por estar em evidências quanto a problemas de agência no âmbito do mercado nacional. Os resultados obtidos apontaram que o mercado reagiu significantemente aos anúncios dos conflitos selecionados, apresentando um valor de p-value <0,05 para os blocos de eventos, o que significa a rejeição da hipótese nula, constatando que a evidencia estatística dos dados testados comprova retornos anormais acumulados diferentes de zero. Entretanto, sugere-se novas pesquisas com outros parâmetros de eventos na busca de mais evidências sobre o efeito das informações no preço das ações.