This research investigates the influence of emotional factors on investment decision-making in Bursa Malaysia during the COVID-19 pandemic, with a focus on Malaysia’s market sentiments. Utilizing big data and advanced machine learning techniques, the study employs sentiment analysis to extract emotional cues from online news. The methodology integrates the use of Python for data analysis, feature selection, and Natural Language Processing (NLP), with a particular emphasis on sentiment analysis models like Finance Bidirectional Encoded Representations from Transformers (FinBERT). This approach allows for an in-depth examination of how investor emotions especially retail investor, as reflected in digital discourse, correlate with stock market trends. The findings of this study reveal a significant relationship between the sentiments of Malaysia’s market and the volatility of stock prices, especially during the pandemic’s peak. The application of sentiment analysis in this context provides a novel perspective on forecasting stock market behaviour, highlighting the often-overlooked psychological elements in financial decision-making. The results demonstrate the potential of sentiment analysis as a tool in understanding market dynamics, particularly in turbulent times.