Our goal is to study how stocks from Mexico and United States are interconnected. We apply a novel method based on a graphical model. We estimate partial correlations for every year of the period 2000-2020. Our results based on partial correlation matrices show a systematic level of inter-connectivity across countries that metrics from network theory confirm. An important difference between these countries is how sectors in each market are linked. Most sector graphs in the United States are densely interconnected. In contrast, sectors in Mexico present much less links. We then compare networks in the periods of the subprime mortgage crisis and the crisis triggered by the COVID-19 pandemic. The different propagation speeds of both crises are correctly captured by the metrics. A limitation is derived from information, and it is desirable to include actualized data in the study. The deployed novel method, which led to obtain new results, endow originality to the work. It is concluded that disaggregated data provides a promising venue of research.

/pdf/a-network-of-two-markets-correlations-for-stocks-in-the-s-m8cyjv62.pdf

A Network of two Markets, Correlations for Stocks in the S&P500 Index and Stocks Traded in the BMV

Our goal in this paper is to study and characterize the interdependency structure of the Mexican Stock Exchange (mainly stocks from Bolsa Mexicana de Valores) for the period 2000-2019 which provide a one shot big-picture panorama. To this end, we estimate correlation/concentration matrices from different models and then compute centralities and modularity from network theory.

/pdf/the-interdependency-structure-in-the-mexican-stock-exchange-4vcz0oski9.pdf

The interdependency structure in the Mexican stock exchange: A network approach.

Our goal in this paper is to study and characterize the interdependency structure of the Mexican Stock Exchange (mainly stocks from BMV) in the period 2000-2019 and provide visualizations which in a one shot provide a big-picture panorama. To this end, we estimate correlation/concentration matrices from different models and then compute metrics from network theory including eigencentralities and network modularity

Geography, for example because of the presence of rivers, ravines, or peaks, can subdivide a city. These subdivisions raise the question of identifying the area occupied by the city, of deciding whether or not they are aggregates of distinct built-up areas, and whether or not geographical separations are merely asperities in a certain continuum of built-up areas. The city as a union of administrative units allows for jurisdictional practices, but for public policy in health for example, identification by built-up areas is more operational. The study of urban populations thus requires that cities be circumscribed on objective criteria. Circumscribing a city requires knowledge of commuting flows, but in the absence of this piece of information, circumscribing it relies on the fact that it is made up of close built-up areas. This is reflected in the intersection of the convex envelopes of the spatial extent of these built-up areas. The algorithm treats coordinates of the vertices of polygons encompassing built-up areas provided by the Census Bureau for the United States or the National Institute of Statistics and Geography for Mexico. It allows for computing whether convex hulls of polygons intersect or not. If they do, then the built-up areas circumscribed by these polygons are part of the same city. The result is that cities now reflect the geographic extent of urban areas rather than their administrative areas. With this delineation method applied to Mexico’s 2020 census data, the population sizes of urban areas with at least 2,126 inhabitants follow a power law, with exponent 0.954 (standard deviation = 0.016), whereas this is no longer the case when considering only the administrative extents of cities with more than 15,000 inhabitants. 

Erick Treviño Aguilar

Papers

The interdependency structure in the Mexican stock exchange: A network approach.

The interdependency structure in the Mexican stock exchange: A network approach.

The population sizes of Mexican cities follow a power-law distribution

A Network of two Markets, Correlations for Stocks in the S&P500 Index and Stocks Traded in the BMV