KeFVP: Knowledge-enhanced Financial Volatility Prediction

Financial volatility prediction is vital for characterizing a company's risk profile. Transcripts of companies' earnings calls serve as valuable, yet unstructured, data sources to be utilized to access companies' performance and risk profiles. Despite their importance, current works ignore the role of financial metrics knowledge (such as EBIT, EPS, and ROI) in transcripts, which is crucial for understanding companies' performance, and little consideration is given to integrating text and price information. In this work, we statistically analyze common financial metrics and create a special dataset centered on these metrics. Then, we propose a knowledge-enhanced financial volatility prediction method (KeFVP) to inject knowledge of financial metrics into text comprehension by knowledge-enhanced adaptive pre-training (KePt) and effectively integrating text and price information by introducing a conditional time series prediction module. Extensive experiments are conducted on three realworld public datasets, and the results indicate that KeFVP is effective and outperforms all the state-of-the-art methods.(1)