Emotional Drivers of Financial Decision-Making: Unveiling the Link between Emotions and Stock Market Behavior (Part 2)

This study is the second part of a three-part analysis (if it meets the review requirements) of emotions carried out by written documents. These documents were collected from eight students who took part in a three-day stock market experiment in January 2025. In the first part of this research (1), which was previously published, a lexical approach was used to analyze the words' emotional weight in the passages of the written documents. In this study, we considered all the passages in order to analyze a single written document for each student. To analyze the emotional charges in the documents, three Artificial Intelligences (ChatGPT, Gemini, and DeepSeek) have been used, and six queries have been selected. Using emotional couples, our results suggest different analytical processes depending on the Artificial Intelligence requested and a lack of uniformity in the emotional couples generated, according to the queries selected. At least, Artificial Intelligences are able to identify a strong primary and basic emotional trend but seem to have trouble capturing more nuanced emotional levels in a consistent manner. Using multiple queries does not improve the consistency of the results, and, indeed, the most generic query leads to the most uniform results among the selected artificial intelligences.