Thesis ESTRUCTURA JERARQUICA DEL MERCADO ACCIONARIO CHILENO
Date
2016
Authors
CARVALLO OSSA, TOMÁS IGNACIO
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
El mapa del mercado accionario chileno se conforma a partir de nodos, cada uno de ellos representa una acción, y por enlaces los cuales unen cada nodo con la que posea la relación más fuerte en el mercado. Esta relación es obtenida a través de un análisis de correlaciones. Este análisis, no permite el diseño del grafo debido a que no satisface los tres axiomas que obliga la métrica Euclideana, por ende, es necesario utilizar una función de distancias para obtener la matriz de distancias D. Posteriormente, se utilizará una metodología de filtración llamada árbol de expansión mínima, que nace a partir de la teoría de grafos, la cual entregará las distancias mínimas entre cada nodo (relación más fuerte) denominada distancia ultra métrica. A partir de ellas, se confeccionará el grafo relacionando cada nodo con el que corresponda, todas estas conexiones se realizan con la ayuda de un programa llamado "Pajek" que, finalmente, entrega el grafo de mercado accionario chileno y con él su respectiva estructura jerárquica.
The Chilean stock market map is made up of nodes, each representing an action, and by links which link each node with the one with the strongest relationship in the market. This relationship is obtained through an analysis of correlations. This analysis does not allow the design of the graph because it does not satisfy the three axioms required by the Euclidean metric, therefore, it is necessary to use a distance function to obtain the distance matrix D. Subsequently, a filtering methodology will be used minimum spanning tree, which is born from the theory of graphs, which will deliver the minimum distances between each node (stronger relationship) called ultrametric distance. From these, the graph will be made relating each node to which it corresponds, all these connections are made with the help of a program called "Pajek" that, finally, delivers the Chilean stock market graph and with it its respective hierarchical structure .
The Chilean stock market map is made up of nodes, each representing an action, and by links which link each node with the one with the strongest relationship in the market. This relationship is obtained through an analysis of correlations. This analysis does not allow the design of the graph because it does not satisfy the three axioms required by the Euclidean metric, therefore, it is necessary to use a distance function to obtain the distance matrix D. Subsequently, a filtering methodology will be used minimum spanning tree, which is born from the theory of graphs, which will deliver the minimum distances between each node (stronger relationship) called ultrametric distance. From these, the graph will be made relating each node to which it corresponds, all these connections are made with the help of a program called "Pajek" that, finally, delivers the Chilean stock market graph and with it its respective hierarchical structure .
Description
Catalogado desde la version PDF de la tesis.
Keywords
ESTRUCTURA JERARQUICA , IGPA , IPSA , MERCADO ACCIONARIO CHILENO