The Banco de la República's announcements are a determining element in the decision-making process of different agents in the Colombian economy, especially banks, institutional investors and international investment funds. The behavior of stock returns listed on the Colombian stock exchange (BVC) offer an approximation to the level of growth, investment, price stability and the general state of the economy. This research analyzes the link between the central bank's announcements and the behavior of the returns of the main stock market indicator and the most dynamic shares listed on the BVC. For this purpose, market sentiment, a concept belonging to the field of Natural Language Processing (NLP), was used. The sources for estimating market sentiment were the minutes of Banco de la República and the monthly returns of COLCAP, and those of Ecopetrol and Bancolombia shares for the period between 2012 and 2022 were considered. Using an ARDL model, the results show a relationship between COLCAP returns and the market sentiment.
