Artificial Intelligence-Based Sentiment Analysis of Indian Union Budget Speeches Using Finbert: Evidence from Selected Indian Information Technology Companies

Abstract

Government budget speeches play a vital role in communicating fiscal priorities, economic reforms, and policy directions that shape business expectations across industries. In recent years, advances in Artificial Intelligence (AI) and Natural Language Processing (NLP) have created new opportunities to evaluate such policy documents using data-driven approaches rather than relying solely on qualitative interpretation. Although several studies have explored sentiment analysis in financial news and corporate disclosures, limited attention has been given to analysing Indian Union Budget speeches using finance-specific language models and examining their relationship with corporate financial performance. This study investigates the sentiment expressed in six Indian Union Budget speeches covering the financial years 2021–22 to 2026–27 and evaluates whether policy sentiment is associated with the financial performance of selected Indian Information Technology (IT) companies. The study employs FinBERT as the primary sentiment-analysis model, while VADER and TextBlob are used as complementary techniques for comparison. Financial performance is assessed using secondary data on revenue, operating profit, and net profit for five leading Indian IT companies: Infosys, Tata Consultancy Services (TCS), Wipro, HCL Technologies, and Tech Mahindra. Pearson correlation analysis is applied to examine the relationship between sentiment scores and financial indicators. The findings indicate that Union Budget speeches generally exhibit a positive policy sentiment across the study period, with noticeable variations in sentiment intensity between financial years. However, correlation analysis suggests that the relationship between budget sentiment and company-level financial performance is generally weak and statistically insignificant for most financial indicators. These results imply that while fiscal policy communication contributes to the broader economic environment, the long-term financial performance of large Indian IT firms is influenced more strongly by international demand, digital transformation, technological innovation, exchange-rate movements, and firm-specific strategic decisions than by domestic budget sentiment alone. The study contributes to the growing literature on AI-enabled financial analysis by demonstrating the application of FinBERT to government policy documents and by integrating sentiment analysis with financial performance evaluation. The findings also provide useful insights for policymakers, financial analysts, corporate managers, and researchers interested in applying AI-based text analytics to economic policy assessment.