Automação de coleta e estruturação de dados da CVM para análise fundamentalista

Abstract

This work presents the development of an automated software for the collection, structuring, and integration of public financial data provided by the Brazilian Securities and Exchange Commission (CVM), focusing on the application of fundamental analysis. The proposed solution aims to overcome the limitations of access, standardization, and organization of financial data disclosed by publicly traded companies, which are traditionally presented in technical and fragmented formats. The software was developed in Python and uses an architecture based on extraction, transformation, and loading stages, supported by a relational model implemented in SQLite, a lightweight, embedded, and easily integrable relational database management system for Python applications. The tool allows the integration of various CVM data categories, such as standardized financial statements, quarterly information, registration forms, and relevant announcements. After processing, these structured data were used to calculate fundamental indicators, demonstrating the practical applicability of the proposed solution. The project contributes to the democratization of financial information, promoting greater transparency, accessibility, and reproducibility of analyses in the financial market. In addition, it supports academic research, the development of quantitative tools, and educational applications in the field of finance.