Project information
Market risk prediction using sentiment analysis

In recent years, critical minerals such as lithium, cobalt, nickel, and uranium have become essential to the global energy transition and technological development. Their growing economic and strategic importance has attracted increasing attention from investors, policymakers, and the media. These markets are characterized by high uncertainty, limited transparency, and strong sensitivity to geopolitical and environmental events, which makes them ideal for exploring the role of sentiment and investor attention in price and risk formation.
The main goal of this project is to improve market risk forecasting by incorporating sentiment and attention measures derived from online and textual data into empirical risk models. The research will focus on identifying sentiment-driven components of volatility and tail risk in critical mineral markets and comparing their impact with traditional financial factors. This will be achieved by collecting large-scale data from diverse online sources (news media, Google Trends, and social networks), applying text mining and machine learning methods to extract sentiment and attention indices, and integrating them into quantile-based and tail risk forecasting models (e.g., VaR and ES).