Using sentiment analysis to study the relationship between subjective expression in financial reports and company...

Area of Science:

• Financial Reporting Analysis
• Text Mining and Natural Language Processing
• Corporate Performance Measurement

Background:

• Non-financial and qualitative information in corporate disclosures increasingly impacts business insights.
• Traditional financial statements, based on past transactions, offer limited forward-looking economic insights.
• Information asymmetry and uncertainty highlight the importance of advanced financial report analysis.

Purpose of the Study:

• To investigate the relationship between subjective expressions within financial reports and overall company performance.
• To quantify subjective sentiment in financial reports using text mining techniques.
• To explore the impact of sentiment on future corporate development.

Main Methods:

• Utilized natural language processing (NLP) and sentiment analysis for text mining of financial reports.
• Quantified subjective expressions and incorporated intermediary, moderating, and control variables.
• Constructed and applied a multiple regression model to analyze the data.

Main Results:

• Empirical evidence demonstrates that underlying emotional tendencies in subjective expressions significantly affect the future development of listed companies.
• The study quantifies the impact of sentiment on corporate performance.
• Identified a multi-dimensional relationship between financial report text and company performance.

Conclusions:

• Subjective sentiment in financial reports is a critical factor influencing corporate future development.
• This research enhances the understanding of how textual elements in financial disclosures relate to company performance.
• Provides valuable insights for stakeholders to make rational decisions, ensuring operational stability and capital appreciation.