Clustering of Assets and its relationship with macroeconomic variables and the financial indexes

• Autor(a) principal: Santos, Daiane Rodrigues dos
• Data de Publicação: 2021
• Outros Autores: Silva, Tuany Esthefany Barcellos de Carvalho, Villagrán, Campo Elias Suárez, Ribeiro, Tiago Costa
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Research, Society and Development
• Texto Completo: https://rsdjournal.org/index.php/rsd/article/view/12421
• topic: Financial assets; Cluster; Financial decisions.; Activos financieros; Racimo; Decisiones financieras.; Ativos financeiros; Cluster; Decisões financeiras.

The advent of the financial market is one of the most fascinating events of our time. Over the years, researchers and investors have been interested in developing tools to assist in making decisions regarding capital allocation. This work proposes clustering as a metric to classify a set of assets, through a grouping method that maximizes similarity between elements of the same group and minimizes similarity between different groups in order to mitigate portfolio risk.  Additionally, we use multiple linear regressions (MLR) to show whether the assets belonging to the clusters respond in a similar way to some macroeconomic variables and financial indexes.  For the analyzed period -January 2019 to January 2020 - we obtained 8 different clusters of assets with a minimum of 1 asset (1.32 % of total assets) and a maximum of 30 assets (42.86% of total assets). With respect to the relationships with the selected variables, the ANBIMA market index (IMAB) and the smll index (smalll caps) are the most closely variables related to the clusters whereas that the IPCA and Ibovespa indexes are the least significant variables in the econometric application proposed in this paper.