Behavioral Economics and Stock Market Sentiments in Investment Decisions in Mexico: Web Scraping, Natural Language Processing, and Pearson Correlation of Scores

Abstract

This paper analyzes market sentiment using digital news to understand the influence of cognitive biases (behavioral economics) on the investment decisions of economic agents. Web scraping techniques were employed to gather news about “América Movil” stock, the company with the highest market capitalization value. Subsequently, Natural Language Processing (NLP) tools are used to determine sentiment polarity scores, Pearson correlation sentiment scores are carried out. The data consists of digital news articles related to América Movil and its corresponding historical stock price data, providing a basis for sentiment analysis and trend comparison. The findings expose a consistent trend between sentiment polarity scores and stock price movements. Moreover, economic and political factors significantly influence sentiment serving as early indicators of stock price behavior. The model has practical implications for behavioral economics, demonstrating how news profoundly influences investment decisions and shapes the behavior of the Mexican market. This research uniquely combines sentiment analysis of digital news with stock price data to highlight the impact of cognitive biases on investment decisions in the Mexican market.