Do AI-Based Approaches Enhance Financial Statement Analysis in Banking and Investment Firms?

Abstract

This thesis investigates whether artificial intelligence (AI)-based approaches enhance financial statement analysis (FSA) in banking and investment firms. By comparing traditional ratio-based analysis with AI-driven evaluations of narrative disclosures, the study assesses the added value of AI in financial decision-making. Two major European banks, BNP Paribas and ING Group, serve as case studies to highlight the complementary strengths of traditional and AI-assisted methods.

The findings reveal that AI significantly improves analytical efficiency and depth, particularly when processing large volumes of unstructured data such as sustainability reports, risk factors, and management discussions. AI proves especially useful in identifying forward-looking insights and qualitative risk signals often overlooked by traditional approaches.

However, the study also underscores key limitations: model transparency (black-box risk), potential data bias, and the critical need for human oversight. Rather than replacing human analysts, AI is most effective when integrated into a hybrid, human-in-the-loop framework.

In conclusion, AI-based tools can enhance but not replace traditional financial analysis. When combined strategically, they create a more robust, forward-looking, and efficient analytical process suited to the growing complexity of financial disclosures.