Explainable Artificial Intelligence for identifying profitability predictors in Financial Statements

2025-01-29

AI Summary

Traditional financial profitability forecasting models suffer from poor interpretability, failing to meet the transparency requirements of the EU AI Act. Method: This study proposes a game-theoretic eXplainable AI (XAI)-driven machine learning framework that directly models raw financial statements of Italian listed companies from 2013–2022. It is the first work to systematically integrate Shapley value theory across the entire profitability trend forecasting pipeline—combining XGBoost, LightGBM, and deep neural networks—and augmenting them with SHAP and LIME for dynamic feature attribution. Leveraging the AIDA database interface and automated feature engineering, the framework identifies five key dynamic drivers, including accounts receivable turnover ratio and gross margin change rate. Contribution/Results: Experiments demonstrate a 12.6% improvement in F1-score, achieving simultaneous advances in predictive accuracy and regulatory compliance with EU AI transparency mandates.

Technology Category

• Machine Learning: Transparent, Interpretable, Explainable ML
• Humans and AI: Explainable AI (XAI) for Human Understanding
• Philosophy and Ethics of AI: Accountability, Interpretability & Explainability

Application Category

• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• User Modeling, Personalization and Recommendation: Accountability, Transparency, and Ethics for personalization

Abstract

The interconnected nature of the economic variables influencing a firm's performance makes the prediction of a company's earning trend a challenging task. Existing methodologies often rely on simplistic models and financial ratios failing to capture the complexity of interacting influences. In this paper, we apply Machine Learning techniques to raw financial statements data taken from AIDA, a Database comprising Italian listed companies' data from 2013 to 2022. We present a comparative study of different models and following the European AI regulations, we complement our analysis by applying explainability techniques to the proposed models. In particular, we propose adopting an eXplainable Artificial Intelligence method based on Game Theory to identify the most sensitive features and make the result more interpretable.

Problem

• Artificial Intelligence
• Financial Statements
• Profitability Prediction

Innovation

• Machine Learning
• Interpretable AI
• Financial Predictive Analytics