Search

El objetivo de esta tesis consiste en analizar cómo se propaga la percepción idiosincrática del riesgo político en las compañías pertenecientes a una gran red de empresas representativa de los distintos sectores que conforman la economía de Estados Unidos. Se utilizan dos métodos distintos para determinar la existencia de transmisión de riesgo político entre las distintas empresas. En primer lugar, se utiliza un algoritmo basado en Lasso, el cual recibe el nombre de Nets, para estimar un modelo VAR grande y escaso, cuyos coeficientes establecen la transmisión de riesgo político entre las empresas. En segundo lugar, se realizan tests de causalidad en el sentido de Granger para determinar la existencia de transmisión de riesgo político entre las 1.099 empresas de la red. Finalmente, esta red se caracteriza mediante distintos estadísticos de riesgo, los cuales incluyen el número de conexiones entrantes, salientes y totales de cada empresa y la centralidad del vector propio.

It is known that in the last years, the number of women enrolled in the Mathematics Degree at UPC has decreased considerably. Based on that, and on the importance to show to society that both men and women are equally able to pursue the same fields of academia, profession, sports etc, the objective of this study is to compare the performance of the students enrolled in the Bachelor's in Mathematics at UPC from a gender point of view. The examined dataset consists of 725 students that were enrolled in the Bachelor's in Mathematics at UPC during the school years 2009/10 to 2019/20. The response variables considered in this study are the bachelor's access grade, the results of the level test applied to the students when they are admitted, the mean grade of the initial phase, the number of semesters required to pass the initial phase, number of subjects passed on the first attempt and the dropout probability. Gender, if the student is or not CFIS, the bachelor's access grade and the lowest access grade, the results of the level test and the year of enrolment have been used as possible explanatory variables. The methodologies that have been applied are Linear Models, Generalized Linear Models with a Binomial response and the analysis of contingency tables.